.gitignore

@@ -9,3 +9,4 @@ pyrightconfig.json
 **/.speakeasy/temp/
 **/.speakeasy/logs/
 .speakeasy/reports
+.env

.speakeasy/OAS_files/overlay.yaml

@@ -1,90 +0,0 @@
-overlay: 1.0.0
-x-speakeasy-jsonpath: rfc9535
-info:
-  title: Overlay chat-completions-openapi.yaml => openapi.yaml
-  version: 0.0.0
-actions:
-  - target: $["components"]["schemas"]["ChatCompletion"]["properties"]["system_fingerprint"]
-    update:
-      type: string
-      nullable: true
-      description: System fingerprint
-  - target: $["components"]["schemas"]["ChatCompletionChunk"]["properties"]["system_fingerprint"]
-    update:
-      type: string
-      nullable: true
-  - target: $["components"]["schemas"]
-    update:
-      ChatStreamCompletionCreateParams:
-        allOf:
-          - $ref: "#/components/schemas/ChatCompletionCreateParams"
-          - type: object
-            properties:
-              stream:
-                type: boolean
-                enum:
-                  - true
-                default: true
-                description: Enable streaming response
-  - target: $["paths"]["/chat/completions"]["post"]["responses"]["200"]["content"]["application/json"]["schema"]
-    update:
-      $ref: '#/components/schemas/ChatCompletion'
-  - target: $["paths"]["/chat/completions"]["post"]["responses"]["200"]["content"]["application/json"]["schema"]
-    update:
-      description: Non-streaming response when stream=false
-  - target: $["paths"]["/chat/completions"]["post"]["responses"]["200"]["content"]["application/json"]["schema"]["oneOf"]
-    remove: true
-  - target: $["paths"]["/chat/completions"]["post"]["responses"]["200"]["content"]["text/event-stream"]
-    remove: true
-  - target: $["paths"]
-    update:
-      /chat/completions#stream:
-        post:
-          operationId: streamChatCompletion
-          summary: Create a chat completion
-          description: Creates a model response for the given chat conversation. Supports both streaming and non-streaming modes.
-          tags:
-            - Chat
-          requestBody:
-            description: Chat completion request parameters
-            required: true
-            content:
-              application/json:
-                schema:
-                  $ref: '#/components/schemas/ChatStreamCompletionCreateParams'
-          responses:
-            '200':
-              description: Successful chat completion response
-              content:
-                text/event-stream:
-                  x-speakeasy-sse-sentinel: '[DONE]'
-                  schema:
-                    type: object
-                    required: [data]
-                    properties:
-                      data:
-                        $ref: '#/components/schemas/ChatCompletionChunk'
-            '400':
-              description: Bad request - invalid parameters
-              content:
-                application/json:
-                  schema:
-                    $ref: '#/components/schemas/ChatCompletionError'
-            '401':
-              description: Unauthorized - invalid API key
-              content:
-                application/json:
-                  schema:
-                    $ref: '#/components/schemas/ChatCompletionError'
-            '429':
-              description: Too many requests - rate limit exceeded
-              content:
-                application/json:
-                  schema:
-                    $ref: '#/components/schemas/ChatCompletionError'
-            '500':
-              description: Internal server error
-              content:
-                application/json:
-                  schema:
-                    $ref: '#/components/schemas/ChatCompletionError'

.speakeasy/OAS_files/speakeasy-modifications-overlay.yaml

@@ -1,28 +0,0 @@
-overlay: 1.0.0
-x-speakeasy-jsonpath: rfc9535
-info:
-  title: Speakeasy Modifications
-  version: 0.0.2
-  x-speakeasy-metadata:
-    after: ""
-    before: ""
-    type: speakeasy-modifications
-actions:
-  - target: $["paths"]["/chat/completions"]["post"]
-    update:
-      x-speakeasy-name-override: complete
-    x-speakeasy-metadata:
-      after: sdk.chat.complete()
-      before: sdk.Chat.createChatCompletion()
-      created_at: 1755805257797
-      reviewed_at: 1755805262583
-      type: method-name
-  - target: $["paths"]["/chat/completions#stream"]["post"]
-    update:
-      x-speakeasy-name-override: completeStream
-    x-speakeasy-metadata:
-      after: sdk.chat.completeStream()
-      before: sdk.Chat.streamChatCompletion()
-      created_at: 1755805257798
-      reviewed_at: 1755805262583
-      type: method-name

.speakeasy/gen.lock

@@ -1,19 +1,20 @@
 lockVersion: 2.0.0
 id: 232c6d4f-b0fd-4172-8f1b-e2421566e9b4
 management:
-  docChecksum: 003bbbc167e4db8bd3f147793a6648e1
+  docChecksum: 98b4f227e767bb91333b3b3469d93fa5
   docVersion: 1.0.0
-  speakeasyVersion: 1.606.2
+  speakeasyVersion: 1.611.1
-  generationVersion: 2.687.1
+  generationVersion: 2.694.1
-  releaseVersion: 0.1.3
+  releaseVersion: 0.5.2
-  configChecksum: d572a56cfb69d7061ad0319f157ff5d8
+  configChecksum: 4619605fbeb6f4488feb5336aa62bab6
   repoURL: https://github.com/speakeasy-sdks/openrouter-python-sdk.git
   installationURL: https://github.com/speakeasy-sdks/openrouter-python-sdk.git
 features:
   python:
     additionalDependencies: 1.0.0
     constsAndDefaults: 1.0.5
-    core: 5.19.9
+    core: 5.20.3
+    customCodeRegions: 0.1.1
     defaultEnabledRetries: 0.2.0
     devContainers: 3.0.0
     enumUnions: 0.1.0
@@ -29,7 +30,7 @@ features:
     responseFormat: 1.0.1
     retries: 3.0.2
     sdkHooks: 1.1.0
-    serverEvents: 1.0.7
+    serverEvents: 1.0.8
     serverEventsSentinels: 0.1.0
     unions: 3.0.4
 generatedFiles:
@@ -308,16 +309,16 @@ examples:
         "200":
           application/json: {"id": "<id>", "choices": [], "created": 6977.95, "model": "El Camino", "object": "chat.completion"}
         "400":
-          application/json: {"error": {"code": "<value>", "message": "<value>", "param": "<value>", "type": "<value>"}}
+          application/json: {"error": {"code": null, "message": "<value>"}}
         "500":
-          application/json: {"error": {"code": "<value>", "message": "<value>", "param": "<value>", "type": "<value>"}}
+          application/json: {"error": {"code": null, "message": "<value>"}}
   streamChatCompletion:
     speakeasy-default-stream-chat-completion:
       requestBody:
         application/json: {"messages": [{"role": "user", "content": "Hello, how are you?"}], "stream": true, "temperature": 1, "top_p": 1}
       responses:
         "400":
-          application/json: {"error": {"code": "<value>", "message": "<value>", "param": "<value>", "type": "<value>"}}
+          application/json: {"error": {"code": null, "message": "<value>"}}
         "500":
-          application/json: {"error": {"code": "<value>", "message": "<value>", "param": "<value>", "type": "<value>"}}
+          application/json: {"error": {"code": null, "message": "<value>"}}
 examplesVersion: 1.0.2

.speakeasy/gen.yaml

@@ -25,18 +25,21 @@ generation:
     generateNewTests: true
     skipResponseBodyAssertions: false
 python:
-  version: 0.1.3
+  version: 0.5.2
   additionalDependencies:
     dev: {}
     main: {}
+  allowedRedefinedBuiltins:
+    - id
+    - object
   authors:
     - Speakeasy
   baseErrorName: OpenRouterError
   clientServerStatusCodesAsErrors: true
   defaultErrorName: OpenRouterDefaultError
   description: Python Client SDK Generated by Speakeasy.
-  enableCustomCodeRegions: false
+  enableCustomCodeRegions: true
-  enumFormat: enum
+  enumFormat: union
   envVarPrefix: OPENROUTER
   fixFlags:
     responseRequiredSep2024: true

.speakeasy/original.openapi.yaml

@@ -1545,4 +1545,4 @@ paths:
           content:
             application/json:
               schema:
-                $ref: '#/components/schemas/ChatCompletionError'
+                $ref: '#/components/schemas/ChatCompletionError'

.speakeasy/out.openapi.yaml

@@ -17,7 +17,7 @@ servers:
         default: openrouter.ai
     description: Production server
 security:
-  - BearerAuth: []
+  - ApiKey: []
 tags:
   - name: Chat
     description: Chat completion operations
@@ -26,11 +26,17 @@ externalDocs:
   url: https://openrouter.ai/docs
 components:
   securitySchemes:
-    BearerAuth:
+    ApiKey:
       type: http
       scheme: bearer
       description: API key as bearer token in Authorization header
   schemas:
+    ChatCompletionChunkWrapper:
+      type: object
+      required: [data]
+      properties:
+        data:
+          $ref: '#/components/schemas/ChatCompletionChunk'
     ChatCompletionMessageToolCall:
       type: object
       properties:
@@ -450,7 +456,7 @@ components:
           type: object
           properties:
             code:
-              type: string
+              type: number
               nullable: true
             message:
               type: string
@@ -462,8 +468,6 @@ components:
           required:
             - code
             - message
-            - param
-            - type
           description: Error object structure
       required:
         - error
@@ -1057,7 +1061,8 @@ components:
           type: string
           description: Unique user identifier
         models:
-          x-speakeasy-name-override: models_llm
+          # Needed to avoid conflict with the `models` directory
+          x-speakeasy-name-override: fallback_models
           type: array
           nullable: true
           items:
@@ -1082,15 +1087,17 @@ components:
               description: >
                 Whether to allow backup providers to serve requests
 
-                - true: (default) when the primary provider (or your custom providers in "order") is unavailable, use the next best provider.
+                - true: (default) when the primary provider (or your custom providers in "order") is unavailable, use
+                the next best provider.
 
                 - false: use only the primary/custom provider, and return the upstream error if it's unavailable.
-
             require_parameters:
               type: boolean
               nullable: true
               description: >-
-                Whether to filter providers to only those that support the parameters you've provided. If this setting is omitted or set to false, then providers will receive only the parameters they support, and ignore the rest.
+                Whether to filter providers to only those that support the parameters you've provided. If this setting
+                is omitted or set to false, then providers will receive only the parameters they support, and ignore the
+                rest.
             data_collection:
               type: string
               nullable: true
@@ -1098,12 +1105,12 @@ components:
                 - deny
                 - allow
               description: >
-                Data collection setting. If no available model provider meets the requirement, your request will return an error.
+                Data collection setting. If no available model provider meets the requirement, your request will return
+                an error.
 
                 - allow: (default) allow providers which store user data non-transiently and may train on it
 
                 - deny: use only providers which do not collect user data.
-
             order:
               type: array
               nullable: true
@@ -1191,7 +1198,9 @@ components:
                       - Z.AI
                   - type: string
               description: >-
-                An ordered list of provider slugs. The router will attempt to use the first provider in the subset of this list that supports your requested model, and fall back to the next if it is unavailable. If no providers are available, the request will fail with an error message.
+                An ordered list of provider slugs. The router will attempt to use the first provider in the subset of
+                this list that supports your requested model, and fall back to the next if it is unavailable. If no
+                providers are available, the request will fail with an error message.
             only:
               type: array
               nullable: true
@@ -1279,7 +1288,8 @@ components:
                       - Z.AI
                   - type: string
               description: >-
-                List of provider slugs to allow. If provided, this list is merged with your account-wide allowed provider settings for this request.
+                List of provider slugs to allow. If provided, this list is merged with your account-wide allowed
+                provider settings for this request.
             ignore:
               type: array
               nullable: true
@@ -1367,7 +1377,8 @@ components:
                       - Z.AI
                   - type: string
               description: >-
-                List of provider slugs to ignore. If provided, this list is merged with your account-wide ignored provider settings for this request.
+                List of provider slugs to ignore. If provided, this list is merged with your account-wide ignored
+                provider settings for this request.
             quantizations:
               type: array
               nullable: true
@@ -1392,7 +1403,8 @@ components:
                 - throughput
                 - latency
               description: >-
-                The sorting strategy to use for this request, if "order" is not specified. When set, no load balancing is performed.
+                The sorting strategy to use for this request, if "order" is not specified. When set, no load balancing
+                is performed.
             max_price:
               type: object
               properties:
@@ -1423,7 +1435,8 @@ components:
                     - {}
               additionalProperties: false
               description: >-
-                The object specifying the maximum price you want to pay for this request. USD price per million tokens, for prompt and completion.
+                The object specifying the maximum price you want to pay for this request. USD price per million tokens,
+                for prompt and completion.
           additionalProperties: false
           description: When multiple model providers are available, optionally indicate your routing preference.
         plugins:
@@ -1488,7 +1501,7 @@ components:
       description: Chat completion request parameters
     ChatStreamCompletionCreateParams:
       allOf:
-        - $ref: "#/components/schemas/ChatCompletionCreateParams"
+        - $ref: '#/components/schemas/ChatCompletionCreateParams'
         - type: object
           properties:
             stream:

.speakeasy/workflow.lock

@@ -1,4 +1,4 @@
-speakeasyVersion: 1.606.2
+speakeasyVersion: 1.611.1
 sources:
     OpenRouter Chat Completions API:
         sourceNamespace: open-router-chat-completions-api
@@ -14,18 +14,13 @@ targets:
         sourceNamespace: open-router-chat-completions-api
         sourceRevisionDigest: sha256:30b19a8759608792fb4d27f6bf28d5982d7e2a8cf5b3faf34ba3507a62f9f106
         sourceBlobDigest: sha256:454f1b880a0882a70f78eacccb2c33e3d9d19134a0738b0e21487984a7bf2e63
-        codeSamplesNamespace: open-router-chat-completions-api-python-code-samples
-        codeSamplesRevisionDigest: sha256:9ce1951254983ff6edc533d67fea93254d9e0d124057ce826721a1449a9ba36c
 workflow:
     workflowVersion: 1.0.0
     speakeasyVersion: latest
     sources:
         OpenRouter Chat Completions API:
             inputs:
-                - location: .speakeasy/OAS_files/chat-completions-openapi.yaml
+                - location: .speakeasy/in.openapi.yaml
-            overlays:
-                - location: .speakeasy/OAS_files/overlay.yaml
-                - location: .speakeasy/OAS_files/speakeasy-modifications-overlay.yaml
             output: .speakeasy/out.openapi.yaml
             registry:
                 location: registry.speakeasyapi.dev/sheldon-vaughn-test/sandbox/open-router-chat-completions-api

.speakeasy/workflow.yaml

@@ -3,10 +3,7 @@ speakeasyVersion: latest
 sources:
     OpenRouter Chat Completions API:
         inputs:
-            - location: .speakeasy/OAS_files/chat-completions-openapi.yaml
+            - location: .speakeasy/in.openapi.yaml
-        overlays:
-            - location: .speakeasy/OAS_files/overlay.yaml
-            - location: .speakeasy/OAS_files/speakeasy-modifications-overlay.yaml
         output: .speakeasy/out.openapi.yaml
         registry:
             location: registry.speakeasyapi.dev/sheldon-vaughn-test/sandbox/open-router-chat-completions-api

README-PYPI.md

@@ -125,17 +125,17 @@ Generally, the SDK will work well with most IDEs out of the box. However, when u
 
 ```python
 # Synchronous Example
-from openrouter import OpenRouter, models
+from openrouter import OpenRouter
 import os
 
 
 with OpenRouter(
-    bearer_auth=os.getenv("OPENROUTER_BEARER_AUTH", ""),
+    api_key=os.getenv("OPENROUTER_API_KEY", ""),
 ) as open_router:
 
     res = open_router.chat.complete(messages=[
         {
-            "role": models.ChatCompletionUserMessageParamRole.USER,
+            "role": "user",
             "content": "Hello, how are you?",
         },
     ], stream=False, temperature=1, top_p=1)
@@ -150,18 +150,18 @@ The same SDK client can also be used to make asynchronous requests by importing
 ```python
 # Asynchronous Example
 import asyncio
-from openrouter import OpenRouter, models
+from openrouter import OpenRouter
 import os
 
 async def main():
 
     async with OpenRouter(
-        bearer_auth=os.getenv("OPENROUTER_BEARER_AUTH", ""),
+        api_key=os.getenv("OPENROUTER_API_KEY", ""),
     ) as open_router:
 
         res = await open_router.chat.complete_async(messages=[
             {
-                "role": models.ChatCompletionUserMessageParamRole.USER,
+                "role": "user",
                 "content": "Hello, how are you?",
             },
         ], stream=False, temperature=1, top_p=1)
@@ -180,23 +180,23 @@ asyncio.run(main())
 
 This SDK supports the following security scheme globally:
 
-| Name          | Type | Scheme      | Environment Variable     |
+| Name      | Type | Scheme      | Environment Variable |
-| ------------- | ---- | ----------- | ------------------------ |
+| --------- | ---- | ----------- | -------------------- |
-| `bearer_auth` | http | HTTP Bearer | `OPENROUTER_BEARER_AUTH` |
+| `api_key` | http | HTTP Bearer | `OPENROUTER_API_KEY` |
 
-To authenticate with the API the `bearer_auth` parameter must be set when initializing the SDK client instance. For example:
+To authenticate with the API the `api_key` parameter must be set when initializing the SDK client instance. For example:
 ```python
-from openrouter import OpenRouter, models
+from openrouter import OpenRouter
 import os
 
 
 with OpenRouter(
-    bearer_auth=os.getenv("OPENROUTER_BEARER_AUTH", ""),
+    api_key=os.getenv("OPENROUTER_API_KEY", ""),
 ) as open_router:
 
     res = open_router.chat.complete(messages=[
         {
-            "role": models.ChatCompletionUserMessageParamRole.USER,
+            "role": "user",
             "content": "Hello, how are you?",
         },
     ], stream=False, temperature=1, top_p=1)
@@ -235,17 +235,17 @@ The stream is also a [Context Manager][context-manager] and can be used with the
 underlying connection when the context is exited.
 
 ```python
-from openrouter import OpenRouter, models
+from openrouter import OpenRouter
 import os
 
 
 with OpenRouter(
-    bearer_auth=os.getenv("OPENROUTER_BEARER_AUTH", ""),
+    api_key=os.getenv("OPENROUTER_API_KEY", ""),
 ) as open_router:
 
     res = open_router.chat.complete_stream(messages=[
         {
-            "role": models.ChatCompletionUserMessageParamRole.USER,
+            "role": "user",
             "content": "Hello, how are you?",
         },
     ], stream=True, temperature=1, top_p=1)
@@ -269,18 +269,18 @@ Some of the endpoints in this SDK support retries. If you use the SDK without an
 
 To change the default retry strategy for a single API call, simply provide a `RetryConfig` object to the call:
 ```python
-from openrouter import OpenRouter, models
+from openrouter import OpenRouter
 from openrouter.utils import BackoffStrategy, RetryConfig
 import os
 
 
 with OpenRouter(
-    bearer_auth=os.getenv("OPENROUTER_BEARER_AUTH", ""),
+    api_key=os.getenv("OPENROUTER_API_KEY", ""),
 ) as open_router:
 
     res = open_router.chat.complete(messages=[
         {
-            "role": models.ChatCompletionUserMessageParamRole.USER,
+            "role": "user",
             "content": "Hello, how are you?",
         },
     ], stream=False, temperature=1, top_p=1,
@@ -293,19 +293,19 @@ with OpenRouter(
 
 If you'd like to override the default retry strategy for all operations that support retries, you can use the `retry_config` optional parameter when initializing the SDK:
 ```python
-from openrouter import OpenRouter, models
+from openrouter import OpenRouter
 from openrouter.utils import BackoffStrategy, RetryConfig
 import os
 
 
 with OpenRouter(
     retry_config=RetryConfig("backoff", BackoffStrategy(1, 50, 1.1, 100), False),
-    bearer_auth=os.getenv("OPENROUTER_BEARER_AUTH", ""),
+    api_key=os.getenv("OPENROUTER_API_KEY", ""),
 ) as open_router:
 
     res = open_router.chat.complete(messages=[
         {
-            "role": models.ChatCompletionUserMessageParamRole.USER,
+            "role": "user",
             "content": "Hello, how are you?",
         },
     ], stream=False, temperature=1, top_p=1)
@@ -332,19 +332,19 @@ with OpenRouter(
 
 ### Example
 ```python
-from openrouter import OpenRouter, errors, models
+from openrouter import OpenRouter, errors
 import os
 
 
 with OpenRouter(
-    bearer_auth=os.getenv("OPENROUTER_BEARER_AUTH", ""),
+    api_key=os.getenv("OPENROUTER_API_KEY", ""),
 ) as open_router:
     res = None
     try:
 
         res = open_router.chat.complete(messages=[
             {
-                "role": models.ChatCompletionUserMessageParamRole.USER,
+                "role": "user",
                 "content": "Hello, how are you?",
             },
         ], stream=False, temperature=1, top_p=1)
@@ -401,18 +401,18 @@ The default server `https://{provider_url}/api/v1` contains variables and is set
 #### Example
 
 ```python
-from openrouter import OpenRouter, models
+from openrouter import OpenRouter
 import os
 
 
 with OpenRouter(
     provider_url="https://ruddy-guacamole.info/"
-    bearer_auth=os.getenv("OPENROUTER_BEARER_AUTH", ""),
+    api_key=os.getenv("OPENROUTER_API_KEY", ""),
 ) as open_router:
 
     res = open_router.chat.complete(messages=[
         {
-            "role": models.ChatCompletionUserMessageParamRole.USER,
+            "role": "user",
             "content": "Hello, how are you?",
         },
     ], stream=False, temperature=1, top_p=1)
@@ -426,18 +426,18 @@ with OpenRouter(
 
 The default server can be overridden globally by passing a URL to the `server_url: str` optional parameter when initializing the SDK client instance. For example:
 ```python
-from openrouter import OpenRouter, models
+from openrouter import OpenRouter
 import os
 
 
 with OpenRouter(
     server_url="https://openrouter.ai/api/v1",
-    bearer_auth=os.getenv("OPENROUTER_BEARER_AUTH", ""),
+    api_key=os.getenv("OPENROUTER_API_KEY", ""),
 ) as open_router:
 
     res = open_router.chat.complete(messages=[
         {
-            "role": models.ChatCompletionUserMessageParamRole.USER,
+            "role": "user",
             "content": "Hello, how are you?",
         },
     ], stream=False, temperature=1, top_p=1)
@@ -542,7 +542,7 @@ import os
 def main():
 
     with OpenRouter(
-        bearer_auth=os.getenv("OPENROUTER_BEARER_AUTH", ""),
+        api_key=os.getenv("OPENROUTER_API_KEY", ""),
     ) as open_router:
         # Rest of application here...
 
@@ -551,7 +551,7 @@ def main():
 async def amain():
 
     async with OpenRouter(
-        bearer_auth=os.getenv("OPENROUTER_BEARER_AUTH", ""),
+        api_key=os.getenv("OPENROUTER_API_KEY", ""),
     ) as open_router:
         # Rest of application here...
 ```

README.md

@@ -125,17 +125,17 @@ Generally, the SDK will work well with most IDEs out of the box. However, when u
 
 ```python
 # Synchronous Example
-from openrouter import OpenRouter, models
+from openrouter import OpenRouter
 import os
 
 
 with OpenRouter(
-    bearer_auth=os.getenv("OPENROUTER_BEARER_AUTH", ""),
+    api_key=os.getenv("OPENROUTER_API_KEY", ""),
 ) as open_router:
 
     res = open_router.chat.complete(messages=[
         {
-            "role": models.ChatCompletionUserMessageParamRole.USER,
+            "role": "user",
             "content": "Hello, how are you?",
         },
     ], stream=False, temperature=1, top_p=1)
@@ -150,18 +150,18 @@ The same SDK client can also be used to make asynchronous requests by importing
 ```python
 # Asynchronous Example
 import asyncio
-from openrouter import OpenRouter, models
+from openrouter import OpenRouter
 import os
 
 async def main():
 
     async with OpenRouter(
-        bearer_auth=os.getenv("OPENROUTER_BEARER_AUTH", ""),
+        api_key=os.getenv("OPENROUTER_API_KEY", ""),
     ) as open_router:
 
         res = await open_router.chat.complete_async(messages=[
             {
-                "role": models.ChatCompletionUserMessageParamRole.USER,
+                "role": "user",
                 "content": "Hello, how are you?",
             },
         ], stream=False, temperature=1, top_p=1)
@@ -180,23 +180,23 @@ asyncio.run(main())
 
 This SDK supports the following security scheme globally:
 
-| Name          | Type | Scheme      | Environment Variable     |
+| Name      | Type | Scheme      | Environment Variable |
-| ------------- | ---- | ----------- | ------------------------ |
+| --------- | ---- | ----------- | -------------------- |
-| `bearer_auth` | http | HTTP Bearer | `OPENROUTER_BEARER_AUTH` |
+| `api_key` | http | HTTP Bearer | `OPENROUTER_API_KEY` |
 
-To authenticate with the API the `bearer_auth` parameter must be set when initializing the SDK client instance. For example:
+To authenticate with the API the `api_key` parameter must be set when initializing the SDK client instance. For example:
 ```python
-from openrouter import OpenRouter, models
+from openrouter import OpenRouter
 import os
 
 
 with OpenRouter(
-    bearer_auth=os.getenv("OPENROUTER_BEARER_AUTH", ""),
+    api_key=os.getenv("OPENROUTER_API_KEY", ""),
 ) as open_router:
 
     res = open_router.chat.complete(messages=[
         {
-            "role": models.ChatCompletionUserMessageParamRole.USER,
+            "role": "user",
             "content": "Hello, how are you?",
         },
     ], stream=False, temperature=1, top_p=1)
@@ -235,17 +235,17 @@ The stream is also a [Context Manager][context-manager] and can be used with the
 underlying connection when the context is exited.
 
 ```python
-from openrouter import OpenRouter, models
+from openrouter import OpenRouter
 import os
 
 
 with OpenRouter(
-    bearer_auth=os.getenv("OPENROUTER_BEARER_AUTH", ""),
+    api_key=os.getenv("OPENROUTER_API_KEY", ""),
 ) as open_router:
 
     res = open_router.chat.complete_stream(messages=[
         {
-            "role": models.ChatCompletionUserMessageParamRole.USER,
+            "role": "user",
             "content": "Hello, how are you?",
         },
     ], stream=True, temperature=1, top_p=1)
@@ -269,18 +269,18 @@ Some of the endpoints in this SDK support retries. If you use the SDK without an
 
 To change the default retry strategy for a single API call, simply provide a `RetryConfig` object to the call:
 ```python
-from openrouter import OpenRouter, models
+from openrouter import OpenRouter
 from openrouter.utils import BackoffStrategy, RetryConfig
 import os
 
 
 with OpenRouter(
-    bearer_auth=os.getenv("OPENROUTER_BEARER_AUTH", ""),
+    api_key=os.getenv("OPENROUTER_API_KEY", ""),
 ) as open_router:
 
     res = open_router.chat.complete(messages=[
         {
-            "role": models.ChatCompletionUserMessageParamRole.USER,
+            "role": "user",
             "content": "Hello, how are you?",
         },
     ], stream=False, temperature=1, top_p=1,
@@ -293,19 +293,19 @@ with OpenRouter(
 
 If you'd like to override the default retry strategy for all operations that support retries, you can use the `retry_config` optional parameter when initializing the SDK:
 ```python
-from openrouter import OpenRouter, models
+from openrouter import OpenRouter
 from openrouter.utils import BackoffStrategy, RetryConfig
 import os
 
 
 with OpenRouter(
     retry_config=RetryConfig("backoff", BackoffStrategy(1, 50, 1.1, 100), False),
-    bearer_auth=os.getenv("OPENROUTER_BEARER_AUTH", ""),
+    api_key=os.getenv("OPENROUTER_API_KEY", ""),
 ) as open_router:
 
     res = open_router.chat.complete(messages=[
         {
-            "role": models.ChatCompletionUserMessageParamRole.USER,
+            "role": "user",
             "content": "Hello, how are you?",
         },
     ], stream=False, temperature=1, top_p=1)
@@ -332,19 +332,19 @@ with OpenRouter(
 
 ### Example
 ```python
-from openrouter import OpenRouter, errors, models
+from openrouter import OpenRouter, errors
 import os
 
 
 with OpenRouter(
-    bearer_auth=os.getenv("OPENROUTER_BEARER_AUTH", ""),
+    api_key=os.getenv("OPENROUTER_API_KEY", ""),
 ) as open_router:
     res = None
     try:
 
         res = open_router.chat.complete(messages=[
             {
-                "role": models.ChatCompletionUserMessageParamRole.USER,
+                "role": "user",
                 "content": "Hello, how are you?",
             },
         ], stream=False, temperature=1, top_p=1)
@@ -401,18 +401,18 @@ The default server `https://{provider_url}/api/v1` contains variables and is set
 #### Example
 
 ```python
-from openrouter import OpenRouter, models
+from openrouter import OpenRouter
 import os
 
 
 with OpenRouter(
     provider_url="https://ruddy-guacamole.info/"
-    bearer_auth=os.getenv("OPENROUTER_BEARER_AUTH", ""),
+    api_key=os.getenv("OPENROUTER_API_KEY", ""),
 ) as open_router:
 
     res = open_router.chat.complete(messages=[
         {
-            "role": models.ChatCompletionUserMessageParamRole.USER,
+            "role": "user",
             "content": "Hello, how are you?",
         },
     ], stream=False, temperature=1, top_p=1)
@@ -426,18 +426,18 @@ with OpenRouter(
 
 The default server can be overridden globally by passing a URL to the `server_url: str` optional parameter when initializing the SDK client instance. For example:
 ```python
-from openrouter import OpenRouter, models
+from openrouter import OpenRouter
 import os
 
 
 with OpenRouter(
     server_url="https://openrouter.ai/api/v1",
-    bearer_auth=os.getenv("OPENROUTER_BEARER_AUTH", ""),
+    api_key=os.getenv("OPENROUTER_API_KEY", ""),
 ) as open_router:
 
     res = open_router.chat.complete(messages=[
         {
-            "role": models.ChatCompletionUserMessageParamRole.USER,
+            "role": "user",
             "content": "Hello, how are you?",
         },
     ], stream=False, temperature=1, top_p=1)
@@ -542,7 +542,7 @@ import os
 def main():
 
     with OpenRouter(
-        bearer_auth=os.getenv("OPENROUTER_BEARER_AUTH", ""),
+        api_key=os.getenv("OPENROUTER_API_KEY", ""),
     ) as open_router:
         # Rest of application here...
 
@@ -551,7 +551,7 @@ def main():
 async def amain():
 
     async with OpenRouter(
-        bearer_auth=os.getenv("OPENROUTER_BEARER_AUTH", ""),
+        api_key=os.getenv("OPENROUTER_API_KEY", ""),
     ) as open_router:
         # Rest of application here...
 ```

USAGE.md

@@ -1,17 +1,17 @@
 <!-- Start SDK Example Usage [usage] -->
 ```python
 # Synchronous Example
-from openrouter import OpenRouter, models
+from openrouter import OpenRouter
 import os
 
 
 with OpenRouter(
-    bearer_auth=os.getenv("OPENROUTER_BEARER_AUTH", ""),
+    api_key=os.getenv("OPENROUTER_API_KEY", ""),
 ) as open_router:
 
     res = open_router.chat.complete(messages=[
         {
-            "role": models.ChatCompletionUserMessageParamRole.USER,
+            "role": "user",
             "content": "Hello, how are you?",
         },
     ], stream=False, temperature=1, top_p=1)
@@ -26,18 +26,18 @@ The same SDK client can also be used to make asynchronous requests by importing
 ```python
 # Asynchronous Example
 import asyncio
-from openrouter import OpenRouter, models
+from openrouter import OpenRouter
 import os
 
 async def main():
 
     async with OpenRouter(
-        bearer_auth=os.getenv("OPENROUTER_BEARER_AUTH", ""),
+        api_key=os.getenv("OPENROUTER_API_KEY", ""),
     ) as open_router:
 
         res = await open_router.chat.complete_async(messages=[
             {
-                "role": models.ChatCompletionUserMessageParamRole.USER,
+                "role": "user",
                 "content": "Hello, how are you?",
             },
         ], stream=False, temperature=1, top_p=1)

docs/models/chatcompletioncreateparams.md

@@ -28,7 +28,7 @@ Chat completion request parameters
 | `tools`                                                                                                                      | List[[models.ChatCompletionTool](../models/chatcompletiontool.md)]                                                           | :heavy_minus_sign:                                                                                                           | Available tools for function calling                                                                                         |                                                                                                                              |
 | `top_p`                                                                                                                      | *OptionalNullable[float]*                                                                                                    | :heavy_minus_sign:                                                                                                           | Nucleus sampling parameter (0-1)                                                                                             |                                                                                                                              |
 | `user`                                                                                                                       | *Optional[str]*                                                                                                              | :heavy_minus_sign:                                                                                                           | Unique user identifier                                                                                                       |                                                                                                                              |
-| `models_llm`                                                                                                                 | List[*str*]                                                                                                                  | :heavy_minus_sign:                                                                                                           | Order of models to fallback to for this request                                                                              |                                                                                                                              |
+| `fallback_models`                                                                                                            | List[*str*]                                                                                                                  | :heavy_minus_sign:                                                                                                           | Order of models to fallback to for this request                                                                              |                                                                                                                              |
 | `reasoning_effort`                                                                                                           | [OptionalNullable[models.ChatCompletionCreateParamsReasoningEffort]](../models/chatcompletioncreateparamsreasoningeffort.md) | :heavy_minus_sign:                                                                                                           | Reasoning effort                                                                                                             |                                                                                                                              |
 | `provider`                                                                                                                   | [OptionalNullable[models.ChatCompletionCreateParamsProvider]](../models/chatcompletioncreateparamsprovider.md)               | :heavy_minus_sign:                                                                                                           | When multiple model providers are available, optionally indicate your routing preference.                                    |                                                                                                                              |
 | `plugins`                                                                                                                    | List[[models.ChatCompletionCreateParamsPluginUnion](../models/chatcompletioncreateparamspluginunion.md)]                     | :heavy_minus_sign:                                                                                                           | Plugins you want to enable for this request, including their settings.                                                       |                                                                                                                              |

docs/models/chatstreamcompletioncreateparams.md

@@ -28,7 +28,7 @@ Chat completion request parameters
 | `tools`                                                                                                                                  | List[[models.ChatCompletionTool](../models/chatcompletiontool.md)]                                                                       | :heavy_minus_sign:                                                                                                                       | Available tools for function calling                                                                                                     |                                                                                                                                          |
 | `top_p`                                                                                                                                  | *OptionalNullable[float]*                                                                                                                | :heavy_minus_sign:                                                                                                                       | Nucleus sampling parameter (0-1)                                                                                                         |                                                                                                                                          |
 | `user`                                                                                                                                   | *Optional[str]*                                                                                                                          | :heavy_minus_sign:                                                                                                                       | Unique user identifier                                                                                                                   |                                                                                                                                          |
-| `models_llm`                                                                                                                             | List[*str*]                                                                                                                              | :heavy_minus_sign:                                                                                                                       | Order of models to fallback to for this request                                                                                          |                                                                                                                                          |
+| `fallback_models`                                                                                                                        | List[*str*]                                                                                                                              | :heavy_minus_sign:                                                                                                                       | Order of models to fallback to for this request                                                                                          |                                                                                                                                          |
 | `reasoning_effort`                                                                                                                       | [OptionalNullable[models.ChatStreamCompletionCreateParamsReasoningEffort]](../models/chatstreamcompletioncreateparamsreasoningeffort.md) | :heavy_minus_sign:                                                                                                                       | Reasoning effort                                                                                                                         |                                                                                                                                          |
 | `provider`                                                                                                                               | [OptionalNullable[models.ChatStreamCompletionCreateParamsProvider]](../models/chatstreamcompletioncreateparamsprovider.md)               | :heavy_minus_sign:                                                                                                                       | When multiple model providers are available, optionally indicate your routing preference.                                                |                                                                                                                                          |
 | `plugins`                                                                                                                                | List[[models.ChatStreamCompletionCreateParamsPluginUnion](../models/chatstreamcompletioncreateparamspluginunion.md)]                     | :heavy_minus_sign:                                                                                                                       | Plugins you want to enable for this request, including their settings.                                                                   |                                                                                                                                          |

docs/models/error.md

@@ -5,9 +5,9 @@ Error object structure
 
 ## Fields
 
-| Field              | Type               | Required           | Description        |
+| Field                   | Type                    | Required                | Description             |
-| ------------------ | ------------------ | ------------------ | ------------------ |
+| ----------------------- | ----------------------- | ----------------------- | ----------------------- |
-| `code`             | *Nullable[str]*    | :heavy_check_mark: | N/A                |
+| `code`                  | *Nullable[float]*       | :heavy_check_mark:      | N/A                     |
-| `message`          | *str*              | :heavy_check_mark: | N/A                |
+| `message`               | *str*                   | :heavy_check_mark:      | N/A                     |
-| `param`            | *Nullable[str]*    | :heavy_check_mark: | N/A                |
+| `param`                 | *OptionalNullable[str]* | :heavy_minus_sign:      | N/A                     |
-| `type`             | *str*              | :heavy_check_mark: | N/A                |
+| `type`                  | *Optional[str]*         | :heavy_minus_sign:      | N/A                     |

docs/models/security.md

@@ -5,4 +5,4 @@
 
 | Field              | Type               | Required           | Description        |
 | ------------------ | ------------------ | ------------------ | ------------------ |
-| `bearer_auth`      | *Optional[str]*    | :heavy_minus_sign: | N/A                |
+| `api_key`          | *Optional[str]*    | :heavy_minus_sign: | N/A                |

docs/sdks/chat/README.md

@@ -18,17 +18,17 @@ Creates a model response for the given chat conversation. Supports both streamin
 
 <!-- UsageSnippet language="python" operationID="createChatCompletion" method="post" path="/chat/completions" -->
 ```python
-from openrouter import OpenRouter, models
+from openrouter import OpenRouter
 import os
 
 
 with OpenRouter(
-    bearer_auth=os.getenv("OPENROUTER_BEARER_AUTH", ""),
+    api_key=os.getenv("OPENROUTER_API_KEY", ""),
 ) as open_router:
 
     res = open_router.chat.complete(messages=[
         {
-            "role": models.ChatCompletionUserMessageParamRole.USER,
+            "role": "user",
             "content": "Hello, how are you?",
         },
     ], stream=False, temperature=1, top_p=1)
@@ -63,7 +63,7 @@ with OpenRouter(
 | `tools`                                                                                                                         | List[[models.ChatCompletionTool](../../models/chatcompletiontool.md)]                                                           | :heavy_minus_sign:                                                                                                              | Available tools for function calling                                                                                            |                                                                                                                                 |
 | `top_p`                                                                                                                         | *OptionalNullable[float]*                                                                                                       | :heavy_minus_sign:                                                                                                              | Nucleus sampling parameter (0-1)                                                                                                |                                                                                                                                 |
 | `user`                                                                                                                          | *Optional[str]*                                                                                                                 | :heavy_minus_sign:                                                                                                              | Unique user identifier                                                                                                          |                                                                                                                                 |
-| `models_llm`                                                                                                                    | List[*str*]                                                                                                                     | :heavy_minus_sign:                                                                                                              | Order of models to fallback to for this request                                                                                 |                                                                                                                                 |
+| `fallback_models`                                                                                                               | List[*str*]                                                                                                                     | :heavy_minus_sign:                                                                                                              | Order of models to fallback to for this request                                                                                 |                                                                                                                                 |
 | `reasoning_effort`                                                                                                              | [OptionalNullable[models.ChatCompletionCreateParamsReasoningEffort]](../../models/chatcompletioncreateparamsreasoningeffort.md) | :heavy_minus_sign:                                                                                                              | Reasoning effort                                                                                                                |                                                                                                                                 |
 | `provider`                                                                                                                      | [OptionalNullable[models.ChatCompletionCreateParamsProvider]](../../models/chatcompletioncreateparamsprovider.md)               | :heavy_minus_sign:                                                                                                              | When multiple model providers are available, optionally indicate your routing preference.                                       |                                                                                                                                 |
 | `plugins`                                                                                                                       | List[[models.ChatCompletionCreateParamsPluginUnion](../../models/chatcompletioncreateparamspluginunion.md)]                     | :heavy_minus_sign:                                                                                                              | Plugins you want to enable for this request, including their settings.                                                          |                                                                                                                                 |
@@ -89,17 +89,17 @@ Creates a model response for the given chat conversation. Supports both streamin
 
 <!-- UsageSnippet language="python" operationID="streamChatCompletion" method="post" path="/chat/completions#stream" -->
 ```python
-from openrouter import OpenRouter, models
+from openrouter import OpenRouter
 import os
 
 
 with OpenRouter(
-    bearer_auth=os.getenv("OPENROUTER_BEARER_AUTH", ""),
+    api_key=os.getenv("OPENROUTER_API_KEY", ""),
 ) as open_router:
 
     res = open_router.chat.complete_stream(messages=[
         {
-            "role": models.ChatCompletionUserMessageParamRole.USER,
+            "role": "user",
             "content": "Hello, how are you?",
         },
     ], stream=True, temperature=1, top_p=1)
@@ -136,7 +136,7 @@ with OpenRouter(
 | `tools`                                                                                                                                     | List[[models.ChatCompletionTool](../../models/chatcompletiontool.md)]                                                                       | :heavy_minus_sign:                                                                                                                          | Available tools for function calling                                                                                                        |                                                                                                                                             |
 | `top_p`                                                                                                                                     | *OptionalNullable[float]*                                                                                                                   | :heavy_minus_sign:                                                                                                                          | Nucleus sampling parameter (0-1)                                                                                                            |                                                                                                                                             |
 | `user`                                                                                                                                      | *Optional[str]*                                                                                                                             | :heavy_minus_sign:                                                                                                                          | Unique user identifier                                                                                                                      |                                                                                                                                             |
-| `models_llm`                                                                                                                                | List[*str*]                                                                                                                                 | :heavy_minus_sign:                                                                                                                          | Order of models to fallback to for this request                                                                                             |                                                                                                                                             |
+| `fallback_models`                                                                                                                           | List[*str*]                                                                                                                                 | :heavy_minus_sign:                                                                                                                          | Order of models to fallback to for this request                                                                                             |                                                                                                                                             |
 | `reasoning_effort`                                                                                                                          | [OptionalNullable[models.ChatStreamCompletionCreateParamsReasoningEffort]](../../models/chatstreamcompletioncreateparamsreasoningeffort.md) | :heavy_minus_sign:                                                                                                                          | Reasoning effort                                                                                                                            |                                                                                                                                             |
 | `provider`                                                                                                                                  | [OptionalNullable[models.ChatStreamCompletionCreateParamsProvider]](../../models/chatstreamcompletioncreateparamsprovider.md)               | :heavy_minus_sign:                                                                                                                          | When multiple model providers are available, optionally indicate your routing preference.                                                   |                                                                                                                                             |
 | `plugins`                                                                                                                                   | List[[models.ChatStreamCompletionCreateParamsPluginUnion](../../models/chatstreamcompletioncreateparamspluginunion.md)]                     | :heavy_minus_sign:                                                                                                                          | Plugins you want to enable for this request, including their settings.                                                                      |                                                                                                                                             |

examples/.env

@@ -1 +0,0 @@
-OPENROUTER_API_KEY="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"

examples/asyncChatCompletion.py

@@ -0,0 +1,24 @@
+dependencies = [
+    "openrouter",
+    "os",
+    "asyncio",
+]
+
+import os
+import asyncio
+from openrouter import OpenRouter
+
+async def main():
+    sdk = OpenRouter(
+        api_key=os.getenv("OPENROUTER_API_KEY"),
+    )
+    result = await sdk.chat.complete_async(
+        messages=[
+            {"role": "user", "content": "Hello, world!"},
+        ],
+        model="openai/gpt-3.5-turbo",
+    )
+    print("Basic async completion:", result)
+
+if __name__ == "__main__":
+    asyncio.run(main())

examples/asyncSseChatCompletion.py

@@ -0,0 +1,32 @@
+dependencies = [
+    "openrouter",
+    "os",
+    "asyncio",
+]
+
+import os
+import asyncio
+from openrouter import OpenRouter
+
+async def main():
+    open_router = OpenRouter(
+        api_key=os.getenv("OPENROUTER_API_KEY", ""),
+    )
+    res = await open_router.chat.complete_async(
+        messages=[
+            {
+                "role": "user",
+                "content": "Hello, how are you?",
+            },
+        ],
+        model="openai/gpt-3.5-turbo",
+        stream=True
+    )
+
+    async for event in res:  # type: ignore
+        if event.data.choices and event.data.choices[0].delta.content:
+            print(event.data.choices[0].delta.content, end='', flush=True)
+    print()
+
+if __name__ == "__main__":
+    asyncio.run(main())

examples/chatCompletion.py

@@ -8,13 +8,12 @@ import os
 from openrouter import OpenRouter
 
 with OpenRouter(
-    bearer_auth=os.getenv("OPENROUTER_API_KEY"),
+    api_key=os.getenv("OPENROUTER_API_KEY"),
 ) as sdk:
     result = sdk.chat.complete(
-        model="openai/gpt-3.5-turbo",
         messages=[
             {"role": "user", "content": "Hello, world!"},
         ],
+        model="openai/gpt-3.5-turbo",
     )
-
+    print("Basic completion:", result)
-print(result)

examples/sseChatCompletion.py

@@ -5,22 +5,24 @@ dependencies = [
 ]
 
 import os
-from openrouter import OpenRouter, models
+from openrouter import OpenRouter
 
 with OpenRouter(
-    bearer_auth=os.getenv("OPENROUTER_API_KEY", ""),
+    api_key=os.getenv("OPENROUTER_API_KEY", ""),
 ) as open_router:
-
+    res = open_router.chat.complete(
-    res = open_router.chat.complete_stream(messages=[
+        messages=[
-        {
+            {
-            "role": "user",
+                "role": "user",
-            "content": "Hello, how are you?",
+                "content": "Hello, how are you?",
-        },
+            },
-        
+        ],
-    ], model="openai/gpt-3.5-turbo", stream=True, temperature=1, top_p=1)
+        model="openai/gpt-3.5-turbo",
+        stream=True
+    )
 
     with res as event_stream:
         for event in event_stream:
             if event.data.choices and event.data.choices[0].delta.content:
                 print(event.data.choices[0].delta.content, end='', flush=True)
-        print()  # New line at the end
+        print()

pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "openrouter"
-version = "0.1.3"
+version = "0.5.2"
 description = "Python Client SDK Generated by Speakeasy."
 authors = [{ name = "Speakeasy" },]
 readme = "README-PYPI.md"
@@ -30,6 +30,7 @@ build-backend = "setuptools.build_meta"
 
 [tool.pytest.ini_options]
 asyncio_default_fixture_loop_scope = "function"
+asyncio_mode = "auto"
 pythonpath = ["src"]
 
 [tool.mypy]

src/openrouter/_version.py

@@ -3,10 +3,10 @@
 import importlib.metadata
 
 __title__: str = "openrouter"
-__version__: str = "0.1.3"
+__version__: str = "0.5.2"
 __openapi_doc_version__: str = "1.0.0"
-__gen_version__: str = "2.687.1"
+__gen_version__: str = "2.694.1"
-__user_agent__: str = "speakeasy-sdk/python 0.1.3 2.687.1 1.0.0 openrouter"
+__user_agent__: str = "speakeasy-sdk/python 0.5.2 2.694.1 1.0.0 openrouter"
 
 try:
     if __package__ is not None:

src/openrouter/basesdk.py

@@ -15,9 +15,19 @@ from urllib.parse import parse_qs, urlparse
 
 class BaseSDK:
     sdk_configuration: SDKConfiguration
+    parent_ref: Optional[object] = None
+    """
+    Reference to the root SDK instance, if any. This will prevent it from
+    being garbage collected while there are active streams.
+    """
 
-    def __init__(self, sdk_config: SDKConfiguration) -> None:
+    def __init__(
+        self,
+        sdk_config: SDKConfiguration,
+        parent_ref: Optional[object] = None,
+    ) -> None:
         self.sdk_configuration = sdk_config
+        self.parent_ref = parent_ref
 
     def _get_url(self, base_url, url_variables):
         sdk_url, sdk_variables = self.sdk_configuration.get_server_details()

src/openrouter/chat.py

@@ -69,7 +69,7 @@ class Chat(BaseSDK):
         ] = None,
         top_p: OptionalNullable[float] = 1,
         user: Optional[str] = None,
-        models_llm: OptionalNullable[List[str]] = UNSET,
+        fallback_models: OptionalNullable[List[str]] = UNSET,
         reasoning_effort: OptionalNullable[
             models.ChatCompletionCreateParamsReasoningEffort
         ] = UNSET,
@@ -115,7 +115,7 @@ class Chat(BaseSDK):
         :param tools: Available tools for function calling
         :param top_p: Nucleus sampling parameter (0-1)
         :param user: Unique user identifier
-        :param models_llm: Order of models to fallback to for this request
+        :param fallback_models: Order of models to fallback to for this request
         :param reasoning_effort: Reasoning effort
         :param provider: When multiple model providers are available, optionally indicate your routing preference.
         :param plugins: Plugins you want to enable for this request, including their settings.
@@ -170,7 +170,7 @@ class Chat(BaseSDK):
             ),
             top_p=top_p,
             user=user,
-            models_llm=models_llm,
+            fallback_models=fallback_models,
             reasoning_effort=reasoning_effort,
             provider=utils.get_pydantic_model(
                 provider, OptionalNullable[models.ChatCompletionCreateParamsProvider]
@@ -305,7 +305,7 @@ class Chat(BaseSDK):
         ] = None,
         top_p: OptionalNullable[float] = 1,
         user: Optional[str] = None,
-        models_llm: OptionalNullable[List[str]] = UNSET,
+        fallback_models: OptionalNullable[List[str]] = UNSET,
         reasoning_effort: OptionalNullable[
             models.ChatCompletionCreateParamsReasoningEffort
         ] = UNSET,
@@ -351,7 +351,7 @@ class Chat(BaseSDK):
         :param tools: Available tools for function calling
         :param top_p: Nucleus sampling parameter (0-1)
         :param user: Unique user identifier
-        :param models_llm: Order of models to fallback to for this request
+        :param fallback_models: Order of models to fallback to for this request
         :param reasoning_effort: Reasoning effort
         :param provider: When multiple model providers are available, optionally indicate your routing preference.
         :param plugins: Plugins you want to enable for this request, including their settings.
@@ -406,7 +406,7 @@ class Chat(BaseSDK):
             ),
             top_p=top_p,
             user=user,
-            models_llm=models_llm,
+            fallback_models=fallback_models,
             reasoning_effort=reasoning_effort,
             provider=utils.get_pydantic_model(
                 provider, OptionalNullable[models.ChatCompletionCreateParamsProvider]
@@ -541,7 +541,7 @@ class Chat(BaseSDK):
         ] = None,
         top_p: OptionalNullable[float] = 1,
         user: Optional[str] = None,
-        models_llm: OptionalNullable[List[str]] = UNSET,
+        fallback_models: OptionalNullable[List[str]] = UNSET,
         reasoning_effort: OptionalNullable[
             models.ChatStreamCompletionCreateParamsReasoningEffort
         ] = UNSET,
@@ -587,7 +587,7 @@ class Chat(BaseSDK):
         :param tools: Available tools for function calling
         :param top_p: Nucleus sampling parameter (0-1)
         :param user: Unique user identifier
-        :param models_llm: Order of models to fallback to for this request
+        :param fallback_models: Order of models to fallback to for this request
         :param reasoning_effort: Reasoning effort
         :param provider: When multiple model providers are available, optionally indicate your routing preference.
         :param plugins: Plugins you want to enable for this request, including their settings.
@@ -643,7 +643,7 @@ class Chat(BaseSDK):
             ),
             top_p=top_p,
             user=user,
-            models_llm=models_llm,
+            fallback_models=fallback_models,
             reasoning_effort=reasoning_effort,
             provider=utils.get_pydantic_model(
                 provider,
@@ -706,6 +706,7 @@ class Chat(BaseSDK):
                     raw, models.StreamChatCompletionResponseBody
                 ),
                 sentinel="[DONE]",
+                client_ref=self,
             )
         if utils.match_response(http_res, ["400", "401", "429"], "application/json"):
             http_res_text = utils.stream_to_text(http_res)
@@ -792,7 +793,7 @@ class Chat(BaseSDK):
         ] = None,
         top_p: OptionalNullable[float] = 1,
         user: Optional[str] = None,
-        models_llm: OptionalNullable[List[str]] = UNSET,
+        fallback_models: OptionalNullable[List[str]] = UNSET,
         reasoning_effort: OptionalNullable[
             models.ChatStreamCompletionCreateParamsReasoningEffort
         ] = UNSET,
@@ -838,7 +839,7 @@ class Chat(BaseSDK):
         :param tools: Available tools for function calling
         :param top_p: Nucleus sampling parameter (0-1)
         :param user: Unique user identifier
-        :param models_llm: Order of models to fallback to for this request
+        :param fallback_models: Order of models to fallback to for this request
         :param reasoning_effort: Reasoning effort
         :param provider: When multiple model providers are available, optionally indicate your routing preference.
         :param plugins: Plugins you want to enable for this request, including their settings.
@@ -894,7 +895,7 @@ class Chat(BaseSDK):
             ),
             top_p=top_p,
             user=user,
-            models_llm=models_llm,
+            fallback_models=fallback_models,
             reasoning_effort=reasoning_effort,
             provider=utils.get_pydantic_model(
                 provider,
@@ -957,6 +958,7 @@ class Chat(BaseSDK):
                     raw, models.StreamChatCompletionResponseBody
                 ),
                 sentinel="[DONE]",
+                client_ref=self,
             )
         if utils.match_response(http_res, ["400", "401", "429"], "application/json"):
             http_res_text = await utils.stream_to_text_async(http_res)

src/openrouter/errors/__init__.py

@@ -1,14 +1,15 @@
 """Code generated by Speakeasy (https://speakeasy.com). DO NOT EDIT."""
 
+from .openroutererror import OpenRouterError
 from typing import TYPE_CHECKING
 from importlib import import_module
 import builtins
+import sys
 
 if TYPE_CHECKING:
     from .chatcompletionerror import ChatCompletionError, ChatCompletionErrorData
     from .no_response_error import NoResponseError
     from .openrouterdefaulterror import OpenRouterDefaultError
-    from .openroutererror import OpenRouterError
     from .responsevalidationerror import ResponseValidationError
 
 __all__ = [
@@ -25,11 +26,22 @@ _dynamic_imports: dict[str, str] = {
     "ChatCompletionErrorData": ".chatcompletionerror",
     "NoResponseError": ".no_response_error",
     "OpenRouterDefaultError": ".openrouterdefaulterror",
-    "OpenRouterError": ".openroutererror",
     "ResponseValidationError": ".responsevalidationerror",
 }
 
 
+def dynamic_import(modname, retries=3):
+    for attempt in range(retries):
+        try:
+            return import_module(modname, __package__)
+        except KeyError:
+            # Clear any half-initialized module and retry
+            sys.modules.pop(modname, None)
+            if attempt == retries - 1:
+                break
+    raise KeyError(f"Failed to import module '{modname}' after {retries} attempts")
+
+
 def __getattr__(attr_name: str) -> object:
     module_name = _dynamic_imports.get(attr_name)
     if module_name is None:
@@ -38,7 +50,7 @@ def __getattr__(attr_name: str) -> object:
         )
 
     try:
-        module = import_module(module_name, __package__)
+        module = dynamic_import(module_name)
         result = getattr(module, attr_name)
         return result
     except ImportError as e:

src/openrouter/models/__init__.py

@@ -3,6 +3,7 @@
 from typing import TYPE_CHECKING
 from importlib import import_module
 import builtins
+import sys
 
 if TYPE_CHECKING:
     from .annotationdetail import AnnotationDetail, AnnotationDetailTypedDict
@@ -931,6 +932,18 @@ _dynamic_imports: dict[str, str] = {
 }
 
 
+def dynamic_import(modname, retries=3):
+    for attempt in range(retries):
+        try:
+            return import_module(modname, __package__)
+        except KeyError:
+            # Clear any half-initialized module and retry
+            sys.modules.pop(modname, None)
+            if attempt == retries - 1:
+                break
+    raise KeyError(f"Failed to import module '{modname}' after {retries} attempts")
+
+
 def __getattr__(attr_name: str) -> object:
     module_name = _dynamic_imports.get(attr_name)
     if module_name is None:
@@ -939,7 +952,7 @@ def __getattr__(attr_name: str) -> object:
         )
 
     try:
-        module = import_module(module_name, __package__)
+        module = dynamic_import(module_name)
         result = getattr(module, attr_name)
         return result
     except ImportError as e:

src/openrouter/models/chatcompletion.py

@@ -3,7 +3,6 @@
 from __future__ import annotations
 from .chatcompletionchoice import ChatCompletionChoice, ChatCompletionChoiceTypedDict
 from .completionusage import CompletionUsage, CompletionUsageTypedDict
-from enum import Enum
 from openrouter.types import (
     BaseModel,
     Nullable,
@@ -12,12 +11,11 @@ from openrouter.types import (
     UNSET_SENTINEL,
 )
 from pydantic import model_serializer
-from typing import List, Optional
+from typing import List, Literal, Optional
 from typing_extensions import NotRequired, TypedDict
 
 
-class ChatCompletionObject(str, Enum):
+ChatCompletionObject = Literal["chat.completion"]
-    CHAT_COMPLETION = "chat.completion"
 
 
 class ChatCompletionTypedDict(TypedDict):

src/openrouter/models/chatcompletionassistantmessageparam.py

@@ -9,7 +9,6 @@ from .chatcompletionmessagetoolcall import (
     ChatCompletionMessageToolCall,
     ChatCompletionMessageToolCallTypedDict,
 )
-from enum import Enum
 from openrouter.types import (
     BaseModel,
     Nullable,
@@ -18,13 +17,11 @@ from openrouter.types import (
     UNSET_SENTINEL,
 )
 from pydantic import model_serializer
-from typing import Any, List, Optional, Union
+from typing import Any, List, Literal, Optional, Union
 from typing_extensions import NotRequired, TypeAliasType, TypedDict
 
 
-class ChatCompletionAssistantMessageParamRole(str, Enum):
+ChatCompletionAssistantMessageParamRole = Literal["assistant"]
-    ASSISTANT = "assistant"
-
 
 ChatCompletionAssistantMessageParamContentTypedDict = TypeAliasType(
     "ChatCompletionAssistantMessageParamContentTypedDict",

src/openrouter/models/chatcompletionchoice.py

@@ -6,7 +6,6 @@ from .chatcompletiontokenlogprobs import (
     ChatCompletionTokenLogprobs,
     ChatCompletionTokenLogprobsTypedDict,
 )
-from enum import Enum
 from openrouter.types import (
     BaseModel,
     Nullable,
@@ -15,17 +14,14 @@ from openrouter.types import (
     UNSET_SENTINEL,
 )
 from pydantic import model_serializer
+from typing import Literal
 from typing_extensions import NotRequired, TypedDict
 
 
-class ChatCompletionChoiceFinishReason(str, Enum):
+ChatCompletionChoiceFinishReason = Literal[
-    r"""Reason the completion finished"""
+    "tool_calls", "stop", "length", "content_filter", "error"
-
+]
-    TOOL_CALLS = "tool_calls"
+r"""Reason the completion finished"""
-    STOP = "stop"
-    LENGTH = "length"
-    CONTENT_FILTER = "content_filter"
-    ERROR = "error"
 
 
 class ChatCompletionChoiceTypedDict(TypedDict):

src/openrouter/models/chatcompletionchunk.py

@@ -6,7 +6,6 @@ from .chatcompletionchunkchoice import (
     ChatCompletionChunkChoiceTypedDict,
 )
 from .completionusage import CompletionUsage, CompletionUsageTypedDict
-from enum import Enum
 from openrouter.types import (
     BaseModel,
     Nullable,
@@ -15,12 +14,11 @@ from openrouter.types import (
     UNSET_SENTINEL,
 )
 from pydantic import model_serializer
-from typing import List, Optional
+from typing import List, Literal, Optional
 from typing_extensions import NotRequired, TypedDict
 
 
-class ChatCompletionChunkObject(str, Enum):
+ChatCompletionChunkObject = Literal["chat.completion.chunk"]
-    CHAT_COMPLETION_CHUNK = "chat.completion.chunk"
 
 
 class ChatCompletionChunkTypedDict(TypedDict):

src/openrouter/models/chatcompletionchunkchoice.py

@@ -9,7 +9,6 @@ from .chatcompletiontokenlogprobs import (
     ChatCompletionTokenLogprobs,
     ChatCompletionTokenLogprobsTypedDict,
 )
-from enum import Enum
 from openrouter.types import (
     BaseModel,
     Nullable,
@@ -18,15 +17,13 @@ from openrouter.types import (
     UNSET_SENTINEL,
 )
 from pydantic import model_serializer
+from typing import Literal
 from typing_extensions import NotRequired, TypedDict
 
 
-class ChatCompletionChunkChoiceFinishReason(str, Enum):
+ChatCompletionChunkChoiceFinishReason = Literal[
-    TOOL_CALLS = "tool_calls"
+    "tool_calls", "stop", "length", "content_filter", "error"
-    STOP = "stop"
+]
-    LENGTH = "length"
-    CONTENT_FILTER = "content_filter"
-    ERROR = "error"
 
 
 class ChatCompletionChunkChoiceTypedDict(TypedDict):

src/openrouter/models/chatcompletionchunkchoicedelta.py

@@ -7,7 +7,6 @@ from .chatcompletionchunkchoicedeltatoolcall import (
     ChatCompletionChunkChoiceDeltaToolCallTypedDict,
 )
 from .reasoningdetail import ReasoningDetail, ReasoningDetailTypedDict
-from enum import Enum
 from openrouter.types import (
     BaseModel,
     Nullable,
@@ -16,14 +15,12 @@ from openrouter.types import (
     UNSET_SENTINEL,
 )
 from pydantic import model_serializer
-from typing import List, Optional
+from typing import List, Literal, Optional
 from typing_extensions import NotRequired, TypedDict
 
 
-class ChatCompletionChunkChoiceDeltaRole(str, Enum):
+ChatCompletionChunkChoiceDeltaRole = Literal["assistant"]
-    r"""The role of the message author"""
+r"""The role of the message author"""
-
-    ASSISTANT = "assistant"
 
 
 class ChatCompletionChunkChoiceDeltaTypedDict(TypedDict):

src/openrouter/models/chatcompletionchunkchoicedeltatoolcall.py

@@ -1,16 +1,13 @@
 """Code generated by Speakeasy (https://speakeasy.com). DO NOT EDIT."""
 
 from __future__ import annotations
-from enum import Enum
 from openrouter.types import BaseModel
-from typing import Optional
+from typing import Literal, Optional
 from typing_extensions import NotRequired, TypedDict
 
 
-class ChatCompletionChunkChoiceDeltaToolCallType(str, Enum):
+ChatCompletionChunkChoiceDeltaToolCallType = Literal["function"]
-    r"""Tool call type"""
+r"""Tool call type"""
-
-    FUNCTION = "function"
 
 
 class ChatCompletionChunkChoiceDeltaToolCallFunctionTypedDict(TypedDict):

src/openrouter/models/chatcompletioncontentpartaudio.py

@@ -1,26 +1,18 @@
 """Code generated by Speakeasy (https://speakeasy.com). DO NOT EDIT."""
 
 from __future__ import annotations
-from enum import Enum
 from openrouter.types import BaseModel
 import pydantic
+from typing import Literal
 from typing_extensions import Annotated, TypedDict
 
 
-class ChatCompletionContentPartAudioType(str, Enum):
+ChatCompletionContentPartAudioType = Literal["input_audio"]
-    INPUT_AUDIO = "input_audio"
 
-
+ChatCompletionContentPartAudioFormat = Literal[
-class ChatCompletionContentPartAudioFormat(str, Enum):
+    "wav", "mp3", "flac", "m4a", "ogg", "pcm16", "pcm24"
-    r"""Audio format"""
+]
-
+r"""Audio format"""
-    WAV = "wav"
-    MP3 = "mp3"
-    FLAC = "flac"
-    M4A = "m4a"
-    OGG = "ogg"
-    PCM16 = "pcm16"
-    PCM24 = "pcm24"
 
 
 class InputAudioTypedDict(TypedDict):

src/openrouter/models/chatcompletioncontentpartimage.py

@@ -1,22 +1,15 @@
 """Code generated by Speakeasy (https://speakeasy.com). DO NOT EDIT."""
 
 from __future__ import annotations
-from enum import Enum
 from openrouter.types import BaseModel
-from typing import Optional
+from typing import Literal, Optional
 from typing_extensions import NotRequired, TypedDict
 
 
-class ChatCompletionContentPartImageType(str, Enum):
+ChatCompletionContentPartImageType = Literal["image_url"]
-    IMAGE_URL = "image_url"
 
-
+Detail = Literal["auto", "low", "high"]
-class Detail(str, Enum):
+r"""Image detail level for vision models"""
-    r"""Image detail level for vision models"""
-
-    AUTO = "auto"
-    LOW = "low"
-    HIGH = "high"
 
 
 class ChatCompletionContentPartImageImageURLTypedDict(TypedDict):

src/openrouter/models/chatcompletioncontentparttext.py

@@ -1,13 +1,12 @@
 """Code generated by Speakeasy (https://speakeasy.com). DO NOT EDIT."""
 
 from __future__ import annotations
-from enum import Enum
 from openrouter.types import BaseModel
+from typing import Literal
 from typing_extensions import TypedDict
 
 
-class ChatCompletionContentPartTextType(str, Enum):
+ChatCompletionContentPartTextType = Literal["text"]
-    TEXT = "text"
 
 
 class ChatCompletionContentPartTextTypedDict(TypedDict):

src/openrouter/models/chatcompletioncreateparams.py

@@ -14,7 +14,6 @@ from .responseformatjsonschemaschema import (
     ResponseFormatJSONSchemaSchema,
     ResponseFormatJSONSchemaSchemaTypedDict,
 )
-from enum import Enum
 from openrouter.types import (
     BaseModel,
     Nullable,
@@ -24,17 +23,12 @@ from openrouter.types import (
 )
 import pydantic
 from pydantic import model_serializer
-from typing import Any, Dict, List, Optional, Union
+from typing import Any, Dict, List, Literal, Optional, Union
 from typing_extensions import Annotated, NotRequired, TypeAliasType, TypedDict
 
 
-class ChatCompletionCreateParamsEffort(str, Enum):
+ChatCompletionCreateParamsEffort = Literal["high", "medium", "low", "minimal"]
-    r"""OpenAI-style reasoning effort setting"""
+r"""OpenAI-style reasoning effort setting"""
-
-    HIGH = "high"
-    MEDIUM = "medium"
-    LOW = "low"
-    MINIMAL = "minimal"
 
 
 class ChatCompletionCreateParamsReasoningTypedDict(TypedDict):
@@ -94,8 +88,7 @@ class ChatCompletionCreateParamsReasoning(BaseModel):
         return m
 
 
-class ChatCompletionCreateParamsTypePython(str, Enum):
+ChatCompletionCreateParamsTypePython = Literal["python"]
-    PYTHON = "python"
 
 
 class ChatCompletionCreateParamsResponseFormatPythonTypedDict(TypedDict):
@@ -110,8 +103,7 @@ class ChatCompletionCreateParamsResponseFormatPython(BaseModel):
     type: ChatCompletionCreateParamsTypePython
 
 
-class ChatCompletionCreateParamsTypeGrammar(str, Enum):
+ChatCompletionCreateParamsTypeGrammar = Literal["grammar"]
-    GRAMMAR = "grammar"
 
 
 class ChatCompletionCreateParamsResponseFormatGrammarTypedDict(TypedDict):
@@ -131,8 +123,7 @@ class ChatCompletionCreateParamsResponseFormatGrammar(BaseModel):
     r"""Custom grammar for text generation"""
 
 
-class ChatCompletionCreateParamsTypeJSONSchema(str, Enum):
+ChatCompletionCreateParamsTypeJSONSchema = Literal["json_schema"]
-    JSON_SCHEMA = "json_schema"
 
 
 class ChatCompletionCreateParamsJSONSchemaTypedDict(TypedDict):
@@ -207,8 +198,7 @@ class ChatCompletionCreateParamsResponseFormatJSONSchema(BaseModel):
     json_schema: ChatCompletionCreateParamsJSONSchema
 
 
-class ChatCompletionCreateParamsTypeJSONObject(str, Enum):
+ChatCompletionCreateParamsTypeJSONObject = Literal["json_object"]
-    JSON_OBJECT = "json_object"
 
 
 class ChatCompletionCreateParamsResponseFormatJSONObjectTypedDict(TypedDict):
@@ -223,8 +213,7 @@ class ChatCompletionCreateParamsResponseFormatJSONObject(BaseModel):
     type: ChatCompletionCreateParamsTypeJSONObject
 
 
-class ChatCompletionCreateParamsTypeText(str, Enum):
+ChatCompletionCreateParamsTypeText = Literal["text"]
-    TEXT = "text"
 
 
 class ChatCompletionCreateParamsResponseFormatTextTypedDict(TypedDict):
@@ -291,106 +280,96 @@ class ChatCompletionCreateParamsStreamOptions(BaseModel):
     r"""Include usage information in streaming response"""
 
 
-class ChatCompletionCreateParamsReasoningEffort(str, Enum):
+ChatCompletionCreateParamsReasoningEffort = Literal["high", "medium", "low", "minimal"]
-    r"""Reasoning effort"""
+r"""Reasoning effort"""
 
-    HIGH = "high"
+ChatCompletionCreateParamsDataCollection = Literal["deny", "allow"]
-    MEDIUM = "medium"
+r"""Data collection setting. If no available model provider meets the requirement, your request will return an error.
-    LOW = "low"
+- allow: (default) allow providers which store user data non-transiently and may train on it
-    MINIMAL = "minimal"
+- deny: use only providers which do not collect user data.
 
+"""
 
-class ChatCompletionCreateParamsDataCollection(str, Enum):
+ChatCompletionCreateParamsOrderEnum = Literal[
-    r"""Data collection setting. If no available model provider meets the requirement, your request will return an error.
+    "AnyScale",
-    - allow: (default) allow providers which store user data non-transiently and may train on it
+    "Cent-ML",
-    - deny: use only providers which do not collect user data.
+    "HuggingFace",
-
+    "Hyperbolic 2",
-    """
+    "Lepton",
-
+    "Lynn 2",
-    DENY = "deny"
+    "Lynn",
-    ALLOW = "allow"
+    "Mancer",
-
+    "Modal",
-
+    "OctoAI",
-class ChatCompletionCreateParamsOrderEnum(str, Enum):
+    "Recursal",
-    ANY_SCALE = "AnyScale"
+    "Reflection",
-    CENT_ML = "Cent-ML"
+    "Replicate",
-    HUGGING_FACE = "HuggingFace"
+    "SambaNova 2",
-    HYPERBOLIC_2 = "Hyperbolic 2"
+    "SF Compute",
-    LEPTON = "Lepton"
+    "Together 2",
-    LYNN_2 = "Lynn 2"
+    "01.AI",
-    LYNN = "Lynn"
+    "AI21",
-    MANCER = "Mancer"
+    "AionLabs",
-    MODAL = "Modal"
+    "Alibaba",
-    OCTO_AI = "OctoAI"
+    "Amazon Bedrock",
-    RECURSAL = "Recursal"
+    "Anthropic",
-    REFLECTION = "Reflection"
+    "AtlasCloud",
-    REPLICATE = "Replicate"
+    "Atoma",
-    SAMBA_NOVA_2 = "SambaNova 2"
+    "Avian",
-    SF_COMPUTE = "SF Compute"
+    "Azure",
-    TOGETHER_2 = "Together 2"
+    "BaseTen",
-    ONE_DOT_AI = "01.AI"
+    "Cerebras",
-    AI21 = "AI21"
+    "Chutes",
-    AION_LABS = "AionLabs"
+    "Cloudflare",
-    ALIBABA = "Alibaba"
+    "Cohere",
-    AMAZON_BEDROCK = "Amazon Bedrock"
+    "CrofAI",
-    ANTHROPIC = "Anthropic"
+    "Crusoe",
-    ATLAS_CLOUD = "AtlasCloud"
+    "DeepInfra",
-    ATOMA = "Atoma"
+    "DeepSeek",
-    AVIAN = "Avian"
+    "Enfer",
-    AZURE = "Azure"
+    "Featherless",
-    BASE_TEN = "BaseTen"
+    "Fireworks",
-    CEREBRAS = "Cerebras"
+    "Friendli",
-    CHUTES = "Chutes"
+    "GMICloud",
-    CLOUDFLARE = "Cloudflare"
+    "Google",
-    COHERE = "Cohere"
+    "Google AI Studio",
-    CROF_AI = "CrofAI"
+    "Groq",
-    CRUSOE = "Crusoe"
+    "Hyperbolic",
-    DEEP_INFRA = "DeepInfra"
+    "Inception",
-    DEEP_SEEK = "DeepSeek"
+    "InferenceNet",
-    ENFER = "Enfer"
+    "Infermatic",
-    FEATHERLESS = "Featherless"
+    "Inflection",
-    FIREWORKS = "Fireworks"
+    "InoCloud",
-    FRIENDLI = "Friendli"
+    "Kluster",
-    GMI_CLOUD = "GMICloud"
+    "Lambda",
-    GOOGLE = "Google"
+    "Liquid",
-    GOOGLE_AI_STUDIO = "Google AI Studio"
+    "Mancer 2",
-    GROQ = "Groq"
+    "Meta",
-    HYPERBOLIC = "Hyperbolic"
+    "Minimax",
-    INCEPTION = "Inception"
+    "Mistral",
-    INFERENCE_NET = "InferenceNet"
+    "Moonshot AI",
-    INFERMATIC = "Infermatic"
+    "Morph",
-    INFLECTION = "Inflection"
+    "NCompass",
-    INO_CLOUD = "InoCloud"
+    "Nebius",
-    KLUSTER = "Kluster"
+    "NextBit",
-    LAMBDA = "Lambda"
+    "Nineteen",
-    LIQUID = "Liquid"
+    "Novita",
-    MANCER_2 = "Mancer 2"
+    "OpenAI",
-    META = "Meta"
+    "OpenInference",
-    MINIMAX = "Minimax"
+    "Parasail",
-    MISTRAL = "Mistral"
+    "Perplexity",
-    MOONSHOT_AI = "Moonshot AI"
+    "Phala",
-    MORPH = "Morph"
+    "SambaNova",
-    N_COMPASS = "NCompass"
+    "Stealth",
-    NEBIUS = "Nebius"
+    "Switchpoint",
-    NEXT_BIT = "NextBit"
+    "Targon",
-    NINETEEN = "Nineteen"
+    "Together",
-    NOVITA = "Novita"
+    "Ubicloud",
-    OPEN_AI = "OpenAI"
+    "Venice",
-    OPEN_INFERENCE = "OpenInference"
+    "WandB",
-    PARASAIL = "Parasail"
+    "xAI",
-    PERPLEXITY = "Perplexity"
+    "Z.AI",
-    PHALA = "Phala"
+]
-    SAMBA_NOVA = "SambaNova"
-    STEALTH = "Stealth"
-    SWITCHPOINT = "Switchpoint"
-    TARGON = "Targon"
-    TOGETHER = "Together"
-    UBICLOUD = "Ubicloud"
-    VENICE = "Venice"
-    WAND_B = "WandB"
-    X_AI = "xAI"
-    Z_AI = "Z.AI"
-
 
 ChatCompletionCreateParamsOrderUnionTypedDict = TypeAliasType(
     "ChatCompletionCreateParamsOrderUnionTypedDict",
@@ -404,86 +383,86 @@ ChatCompletionCreateParamsOrderUnion = TypeAliasType(
 )
 
 
-class ChatCompletionCreateParamsOnlyEnum(str, Enum):
+ChatCompletionCreateParamsOnlyEnum = Literal[
-    ANY_SCALE = "AnyScale"
+    "AnyScale",
-    CENT_ML = "Cent-ML"
+    "Cent-ML",
-    HUGGING_FACE = "HuggingFace"
+    "HuggingFace",
-    HYPERBOLIC_2 = "Hyperbolic 2"
+    "Hyperbolic 2",
-    LEPTON = "Lepton"
+    "Lepton",
-    LYNN_2 = "Lynn 2"
+    "Lynn 2",
-    LYNN = "Lynn"
+    "Lynn",
-    MANCER = "Mancer"
+    "Mancer",
-    MODAL = "Modal"
+    "Modal",
-    OCTO_AI = "OctoAI"
+    "OctoAI",
-    RECURSAL = "Recursal"
+    "Recursal",
-    REFLECTION = "Reflection"
+    "Reflection",
-    REPLICATE = "Replicate"
+    "Replicate",
-    SAMBA_NOVA_2 = "SambaNova 2"
+    "SambaNova 2",
-    SF_COMPUTE = "SF Compute"
+    "SF Compute",
-    TOGETHER_2 = "Together 2"
+    "Together 2",
-    ONE_DOT_AI = "01.AI"
+    "01.AI",
-    AI21 = "AI21"
+    "AI21",
-    AION_LABS = "AionLabs"
+    "AionLabs",
-    ALIBABA = "Alibaba"
+    "Alibaba",
-    AMAZON_BEDROCK = "Amazon Bedrock"
+    "Amazon Bedrock",
-    ANTHROPIC = "Anthropic"
+    "Anthropic",
-    ATLAS_CLOUD = "AtlasCloud"
+    "AtlasCloud",
-    ATOMA = "Atoma"
+    "Atoma",
-    AVIAN = "Avian"
+    "Avian",
-    AZURE = "Azure"
+    "Azure",
-    BASE_TEN = "BaseTen"
+    "BaseTen",
-    CEREBRAS = "Cerebras"
+    "Cerebras",
-    CHUTES = "Chutes"
+    "Chutes",
-    CLOUDFLARE = "Cloudflare"
+    "Cloudflare",
-    COHERE = "Cohere"
+    "Cohere",
-    CROF_AI = "CrofAI"
+    "CrofAI",
-    CRUSOE = "Crusoe"
+    "Crusoe",
-    DEEP_INFRA = "DeepInfra"
+    "DeepInfra",
-    DEEP_SEEK = "DeepSeek"
+    "DeepSeek",
-    ENFER = "Enfer"
+    "Enfer",
-    FEATHERLESS = "Featherless"
+    "Featherless",
-    FIREWORKS = "Fireworks"
+    "Fireworks",
-    FRIENDLI = "Friendli"
+    "Friendli",
-    GMI_CLOUD = "GMICloud"
+    "GMICloud",
-    GOOGLE = "Google"
+    "Google",
-    GOOGLE_AI_STUDIO = "Google AI Studio"
+    "Google AI Studio",
-    GROQ = "Groq"
+    "Groq",
-    HYPERBOLIC = "Hyperbolic"
+    "Hyperbolic",
-    INCEPTION = "Inception"
+    "Inception",
-    INFERENCE_NET = "InferenceNet"
+    "InferenceNet",
-    INFERMATIC = "Infermatic"
+    "Infermatic",
-    INFLECTION = "Inflection"
+    "Inflection",
-    INO_CLOUD = "InoCloud"
+    "InoCloud",
-    KLUSTER = "Kluster"
+    "Kluster",
-    LAMBDA = "Lambda"
+    "Lambda",
-    LIQUID = "Liquid"
+    "Liquid",
-    MANCER_2 = "Mancer 2"
+    "Mancer 2",
-    META = "Meta"
+    "Meta",
-    MINIMAX = "Minimax"
+    "Minimax",
-    MISTRAL = "Mistral"
+    "Mistral",
-    MOONSHOT_AI = "Moonshot AI"
+    "Moonshot AI",
-    MORPH = "Morph"
+    "Morph",
-    N_COMPASS = "NCompass"
+    "NCompass",
-    NEBIUS = "Nebius"
+    "Nebius",
-    NEXT_BIT = "NextBit"
+    "NextBit",
-    NINETEEN = "Nineteen"
+    "Nineteen",
-    NOVITA = "Novita"
+    "Novita",
-    OPEN_AI = "OpenAI"
+    "OpenAI",
-    OPEN_INFERENCE = "OpenInference"
+    "OpenInference",
-    PARASAIL = "Parasail"
+    "Parasail",
-    PERPLEXITY = "Perplexity"
+    "Perplexity",
-    PHALA = "Phala"
+    "Phala",
-    SAMBA_NOVA = "SambaNova"
+    "SambaNova",
-    STEALTH = "Stealth"
+    "Stealth",
-    SWITCHPOINT = "Switchpoint"
+    "Switchpoint",
-    TARGON = "Targon"
+    "Targon",
-    TOGETHER = "Together"
+    "Together",
-    UBICLOUD = "Ubicloud"
+    "Ubicloud",
-    VENICE = "Venice"
+    "Venice",
-    WAND_B = "WandB"
+    "WandB",
-    X_AI = "xAI"
+    "xAI",
-    Z_AI = "Z.AI"
+    "Z.AI",
-
+]
 
 ChatCompletionCreateParamsOnlyUnionTypedDict = TypeAliasType(
     "ChatCompletionCreateParamsOnlyUnionTypedDict",
@@ -497,86 +476,86 @@ ChatCompletionCreateParamsOnlyUnion = TypeAliasType(
 )
 
 
-class ChatCompletionCreateParamsIgnoreEnum(str, Enum):
+ChatCompletionCreateParamsIgnoreEnum = Literal[
-    ANY_SCALE = "AnyScale"
+    "AnyScale",
-    CENT_ML = "Cent-ML"
+    "Cent-ML",
-    HUGGING_FACE = "HuggingFace"
+    "HuggingFace",
-    HYPERBOLIC_2 = "Hyperbolic 2"
+    "Hyperbolic 2",
-    LEPTON = "Lepton"
+    "Lepton",
-    LYNN_2 = "Lynn 2"
+    "Lynn 2",
-    LYNN = "Lynn"
+    "Lynn",
-    MANCER = "Mancer"
+    "Mancer",
-    MODAL = "Modal"
+    "Modal",
-    OCTO_AI = "OctoAI"
+    "OctoAI",
-    RECURSAL = "Recursal"
+    "Recursal",
-    REFLECTION = "Reflection"
+    "Reflection",
-    REPLICATE = "Replicate"
+    "Replicate",
-    SAMBA_NOVA_2 = "SambaNova 2"
+    "SambaNova 2",
-    SF_COMPUTE = "SF Compute"
+    "SF Compute",
-    TOGETHER_2 = "Together 2"
+    "Together 2",
-    ONE_DOT_AI = "01.AI"
+    "01.AI",
-    AI21 = "AI21"
+    "AI21",
-    AION_LABS = "AionLabs"
+    "AionLabs",
-    ALIBABA = "Alibaba"
+    "Alibaba",
-    AMAZON_BEDROCK = "Amazon Bedrock"
+    "Amazon Bedrock",
-    ANTHROPIC = "Anthropic"
+    "Anthropic",
-    ATLAS_CLOUD = "AtlasCloud"
+    "AtlasCloud",
-    ATOMA = "Atoma"
+    "Atoma",
-    AVIAN = "Avian"
+    "Avian",
-    AZURE = "Azure"
+    "Azure",
-    BASE_TEN = "BaseTen"
+    "BaseTen",
-    CEREBRAS = "Cerebras"
+    "Cerebras",
-    CHUTES = "Chutes"
+    "Chutes",
-    CLOUDFLARE = "Cloudflare"
+    "Cloudflare",
-    COHERE = "Cohere"
+    "Cohere",
-    CROF_AI = "CrofAI"
+    "CrofAI",
-    CRUSOE = "Crusoe"
+    "Crusoe",
-    DEEP_INFRA = "DeepInfra"
+    "DeepInfra",
-    DEEP_SEEK = "DeepSeek"
+    "DeepSeek",
-    ENFER = "Enfer"
+    "Enfer",
-    FEATHERLESS = "Featherless"
+    "Featherless",
-    FIREWORKS = "Fireworks"
+    "Fireworks",
-    FRIENDLI = "Friendli"
+    "Friendli",
-    GMI_CLOUD = "GMICloud"
+    "GMICloud",
-    GOOGLE = "Google"
+    "Google",
-    GOOGLE_AI_STUDIO = "Google AI Studio"
+    "Google AI Studio",
-    GROQ = "Groq"
+    "Groq",
-    HYPERBOLIC = "Hyperbolic"
+    "Hyperbolic",
-    INCEPTION = "Inception"
+    "Inception",
-    INFERENCE_NET = "InferenceNet"
+    "InferenceNet",
-    INFERMATIC = "Infermatic"
+    "Infermatic",
-    INFLECTION = "Inflection"
+    "Inflection",
-    INO_CLOUD = "InoCloud"
+    "InoCloud",
-    KLUSTER = "Kluster"
+    "Kluster",
-    LAMBDA = "Lambda"
+    "Lambda",
-    LIQUID = "Liquid"
+    "Liquid",
-    MANCER_2 = "Mancer 2"
+    "Mancer 2",
-    META = "Meta"
+    "Meta",
-    MINIMAX = "Minimax"
+    "Minimax",
-    MISTRAL = "Mistral"
+    "Mistral",
-    MOONSHOT_AI = "Moonshot AI"
+    "Moonshot AI",
-    MORPH = "Morph"
+    "Morph",
-    N_COMPASS = "NCompass"
+    "NCompass",
-    NEBIUS = "Nebius"
+    "Nebius",
-    NEXT_BIT = "NextBit"
+    "NextBit",
-    NINETEEN = "Nineteen"
+    "Nineteen",
-    NOVITA = "Novita"
+    "Novita",
-    OPEN_AI = "OpenAI"
+    "OpenAI",
-    OPEN_INFERENCE = "OpenInference"
+    "OpenInference",
-    PARASAIL = "Parasail"
+    "Parasail",
-    PERPLEXITY = "Perplexity"
+    "Perplexity",
-    PHALA = "Phala"
+    "Phala",
-    SAMBA_NOVA = "SambaNova"
+    "SambaNova",
-    STEALTH = "Stealth"
+    "Stealth",
-    SWITCHPOINT = "Switchpoint"
+    "Switchpoint",
-    TARGON = "Targon"
+    "Targon",
-    TOGETHER = "Together"
+    "Together",
-    UBICLOUD = "Ubicloud"
+    "Ubicloud",
-    VENICE = "Venice"
+    "Venice",
-    WAND_B = "WandB"
+    "WandB",
-    X_AI = "xAI"
+    "xAI",
-    Z_AI = "Z.AI"
+    "Z.AI",
-
+]
 
 ChatCompletionCreateParamsIgnoreUnionTypedDict = TypeAliasType(
     "ChatCompletionCreateParamsIgnoreUnionTypedDict",
@@ -590,25 +569,12 @@ ChatCompletionCreateParamsIgnoreUnion = TypeAliasType(
 )
 
 
-class ChatCompletionCreateParamsQuantization(str, Enum):
+ChatCompletionCreateParamsQuantization = Literal[
-    INT4 = "int4"
+    "int4", "int8", "fp4", "fp6", "fp8", "fp16", "bf16", "fp32", "unknown"
-    INT8 = "int8"
+]
-    FP4 = "fp4"
-    FP6 = "fp6"
-    FP8 = "fp8"
-    FP16 = "fp16"
-    BF16 = "bf16"
-    FP32 = "fp32"
-    UNKNOWN = "unknown"
-
-
-class ChatCompletionCreateParamsSort(str, Enum):
-    r"""The sorting strategy to use for this request, if \"order\" is not specified. When set, no load balancing is performed."""
-
-    PRICE = "price"
-    THROUGHPUT = "throughput"
-    LATENCY = "latency"
 
+ChatCompletionCreateParamsSort = Literal["price", "throughput", "latency"]
+r"""The sorting strategy to use for this request, if \"order\" is not specified. When set, no load balancing is performed."""
 
 ChatCompletionCreateParamsPromptTypedDict = TypeAliasType(
     "ChatCompletionCreateParamsPromptTypedDict", Union[float, str, Any]
@@ -805,14 +771,9 @@ class ChatCompletionCreateParamsProvider(BaseModel):
         return m
 
 
-class ChatCompletionCreateParamsIDFileParser(str, Enum):
+ChatCompletionCreateParamsIDFileParser = Literal["file-parser"]
-    FILE_PARSER = "file-parser"
 
-
+ChatCompletionCreateParamsPdfEngine = Literal["mistral-ocr", "pdf-text", "native"]
-class ChatCompletionCreateParamsPdfEngine(str, Enum):
-    MISTRAL_OCR = "mistral-ocr"
-    PDF_TEXT = "pdf-text"
-    NATIVE = "native"
 
 
 class ChatCompletionCreateParamsPdfTypedDict(TypedDict):
@@ -837,8 +798,7 @@ class ChatCompletionCreateParamsPluginFileParser(BaseModel):
     pdf: Optional[ChatCompletionCreateParamsPdf] = None
 
 
-class ChatCompletionCreateParamsIDChainOfThought(str, Enum):
+ChatCompletionCreateParamsIDChainOfThought = Literal["chain-of-thought"]
-    CHAIN_OF_THOUGHT = "chain-of-thought"
 
 
 class ChatCompletionCreateParamsPluginChainOfThoughtTypedDict(TypedDict):
@@ -849,13 +809,9 @@ class ChatCompletionCreateParamsPluginChainOfThought(BaseModel):
     id: ChatCompletionCreateParamsIDChainOfThought
 
 
-class ChatCompletionCreateParamsIDWeb(str, Enum):
+ChatCompletionCreateParamsIDWeb = Literal["web"]
-    WEB = "web"
 
-
+ChatCompletionCreateParamsEngine = Literal["native", "exa"]
-class ChatCompletionCreateParamsEngine(str, Enum):
-    NATIVE = "native"
-    EXA = "exa"
 
 
 class ChatCompletionCreateParamsPluginWebTypedDict(TypedDict):
@@ -875,8 +831,7 @@ class ChatCompletionCreateParamsPluginWeb(BaseModel):
     engine: Optional[ChatCompletionCreateParamsEngine] = None
 
 
-class ChatCompletionCreateParamsIDModeration(str, Enum):
+ChatCompletionCreateParamsIDModeration = Literal["moderation"]
-    MODERATION = "moderation"
 
 
 class ChatCompletionCreateParamsPluginModerationTypedDict(TypedDict):
@@ -955,7 +910,7 @@ class ChatCompletionCreateParamsTypedDict(TypedDict):
     r"""Nucleus sampling parameter (0-1)"""
     user: NotRequired[str]
     r"""Unique user identifier"""
-    models_llm: NotRequired[Nullable[List[str]]]
+    fallback_models: NotRequired[Nullable[List[str]]]
     r"""Order of models to fallback to for this request"""
     reasoning_effort: NotRequired[Nullable[ChatCompletionCreateParamsReasoningEffort]]
     r"""Reasoning effort"""
@@ -1030,7 +985,7 @@ class ChatCompletionCreateParams(BaseModel):
     user: Optional[str] = None
     r"""Unique user identifier"""
 
-    models_llm: Annotated[
+    fallback_models: Annotated[
         OptionalNullable[List[str]], pydantic.Field(alias="models")
     ] = UNSET
     r"""Order of models to fallback to for this request"""
@@ -1069,7 +1024,7 @@ class ChatCompletionCreateParams(BaseModel):
             "tools",
             "top_p",
             "user",
-            "models_llm",
+            "fallback_models",
             "reasoning_effort",
             "provider",
             "plugins",
@@ -1089,7 +1044,7 @@ class ChatCompletionCreateParams(BaseModel):
             "stream_options",
             "temperature",
             "top_p",
-            "models_llm",
+            "fallback_models",
             "reasoning_effort",
             "provider",
         ]

src/openrouter/models/chatcompletionerror.py

@@ -1,34 +1,41 @@
 """Code generated by Speakeasy (https://speakeasy.com). DO NOT EDIT."""
 
 from __future__ import annotations
-from openrouter.types import BaseModel, Nullable, UNSET_SENTINEL
+from openrouter.types import (
+    BaseModel,
+    Nullable,
+    OptionalNullable,
+    UNSET,
+    UNSET_SENTINEL,
+)
 from pydantic import model_serializer
-from typing_extensions import TypedDict
+from typing import Optional
+from typing_extensions import NotRequired, TypedDict
 
 
 class ErrorTypedDict(TypedDict):
     r"""Error object structure"""
 
-    code: Nullable[str]
+    code: Nullable[float]
     message: str
-    param: Nullable[str]
+    param: NotRequired[Nullable[str]]
-    type: str
+    type: NotRequired[str]
 
 
 class Error(BaseModel):
     r"""Error object structure"""
 
-    code: Nullable[str]
+    code: Nullable[float]
 
     message: str
 
-    param: Nullable[str]
+    param: OptionalNullable[str] = UNSET
 
-    type: str
+    type: Optional[str] = None
 
     @model_serializer(mode="wrap")
     def serialize_model(self, handler):
-        optional_fields = []
+        optional_fields = ["param", "type"]
         nullable_fields = ["code", "param"]
         null_default_fields = []
 

src/openrouter/models/chatcompletionmessage.py

@@ -7,7 +7,6 @@ from .chatcompletionmessagetoolcall import (
     ChatCompletionMessageToolCallTypedDict,
 )
 from .reasoningdetail import ReasoningDetail, ReasoningDetailTypedDict
-from enum import Enum
 from openrouter.types import (
     BaseModel,
     Nullable,
@@ -16,12 +15,11 @@ from openrouter.types import (
     UNSET_SENTINEL,
 )
 from pydantic import model_serializer
-from typing import List, Optional
+from typing import List, Literal, Optional
 from typing_extensions import NotRequired, TypedDict
 
 
-class ChatCompletionMessageRole(str, Enum):
+ChatCompletionMessageRole = Literal["assistant"]
-    ASSISTANT = "assistant"
 
 
 class ChatCompletionMessageTypedDict(TypedDict):

src/openrouter/models/chatcompletionmessagetoolcall.py

@@ -1,13 +1,12 @@
 """Code generated by Speakeasy (https://speakeasy.com). DO NOT EDIT."""
 
 from __future__ import annotations
-from enum import Enum
 from openrouter.types import BaseModel
+from typing import Literal
 from typing_extensions import TypedDict
 
 
-class ChatCompletionMessageToolCallType(str, Enum):
+ChatCompletionMessageToolCallType = Literal["function"]
-    FUNCTION = "function"
 
 
 class ChatCompletionMessageToolCallFunctionTypedDict(TypedDict):

src/openrouter/models/chatcompletionnamedtoolchoice.py

@@ -1,13 +1,12 @@
 """Code generated by Speakeasy (https://speakeasy.com). DO NOT EDIT."""
 
 from __future__ import annotations
-from enum import Enum
 from openrouter.types import BaseModel
+from typing import Literal
 from typing_extensions import TypedDict
 
 
-class ChatCompletionNamedToolChoiceType(str, Enum):
+ChatCompletionNamedToolChoiceType = Literal["function"]
-    FUNCTION = "function"
 
 
 class ChatCompletionNamedToolChoiceFunctionTypedDict(TypedDict):

src/openrouter/models/chatcompletionsystemmessageparam.py

@@ -5,15 +5,12 @@ from .chatcompletioncontentparttext import (
     ChatCompletionContentPartText,
     ChatCompletionContentPartTextTypedDict,
 )
-from enum import Enum
 from openrouter.types import BaseModel
-from typing import List, Optional, Union
+from typing import List, Literal, Optional, Union
 from typing_extensions import NotRequired, TypeAliasType, TypedDict
 
 
-class ChatCompletionSystemMessageParamRole(str, Enum):
+ChatCompletionSystemMessageParamRole = Literal["system"]
-    SYSTEM = "system"
-
 
 ChatCompletionSystemMessageParamContentTypedDict = TypeAliasType(
     "ChatCompletionSystemMessageParamContentTypedDict",

src/openrouter/models/chatcompletiontool.py

@@ -1,7 +1,6 @@
 """Code generated by Speakeasy (https://speakeasy.com). DO NOT EDIT."""
 
 from __future__ import annotations
-from enum import Enum
 from openrouter.types import (
     BaseModel,
     Nullable,
@@ -10,12 +9,11 @@ from openrouter.types import (
     UNSET_SENTINEL,
 )
 from pydantic import model_serializer
-from typing import Optional
+from typing import Literal, Optional
 from typing_extensions import NotRequired, TypedDict
 
 
-class ChatCompletionToolType(str, Enum):
+ChatCompletionToolType = Literal["function"]
-    FUNCTION = "function"
 
 
 class ParametersTypedDict(TypedDict):

src/openrouter/models/chatcompletiontoolchoiceoption.py

@@ -5,22 +5,15 @@ from .chatcompletionnamedtoolchoice import (
     ChatCompletionNamedToolChoice,
     ChatCompletionNamedToolChoiceTypedDict,
 )
-from enum import Enum
+from typing import Literal, Union
-from typing import Union
 from typing_extensions import TypeAliasType
 
 
-class ChatCompletionToolChoiceOptionRequired(str, Enum):
+ChatCompletionToolChoiceOptionRequired = Literal["required"]
-    REQUIRED = "required"
 
+ChatCompletionToolChoiceOptionAuto = Literal["auto"]
 
-class ChatCompletionToolChoiceOptionAuto(str, Enum):
+ChatCompletionToolChoiceOptionNone = Literal["none"]
-    AUTO = "auto"
-
-
-class ChatCompletionToolChoiceOptionNone(str, Enum):
-    NONE = "none"
-
 
 ChatCompletionToolChoiceOptionTypedDict = TypeAliasType(
     "ChatCompletionToolChoiceOptionTypedDict",

src/openrouter/models/chatcompletiontoolmessageparam.py

@@ -5,15 +5,12 @@ from .chatcompletioncontentpart import (
     ChatCompletionContentPart,
     ChatCompletionContentPartTypedDict,
 )
-from enum import Enum
 from openrouter.types import BaseModel
-from typing import List, Union
+from typing import List, Literal, Union
 from typing_extensions import TypeAliasType, TypedDict
 
 
-class ChatCompletionToolMessageParamRole(str, Enum):
+ChatCompletionToolMessageParamRole = Literal["tool"]
-    TOOL = "tool"
-
 
 ChatCompletionToolMessageParamContentTypedDict = TypeAliasType(
     "ChatCompletionToolMessageParamContentTypedDict",

src/openrouter/models/chatcompletionusermessageparam.py

@@ -5,15 +5,12 @@ from .chatcompletioncontentpart import (
     ChatCompletionContentPart,
     ChatCompletionContentPartTypedDict,
 )
-from enum import Enum
 from openrouter.types import BaseModel
-from typing import List, Optional, Union
+from typing import List, Literal, Optional, Union
 from typing_extensions import NotRequired, TypeAliasType, TypedDict
 
 
-class ChatCompletionUserMessageParamRole(str, Enum):
+ChatCompletionUserMessageParamRole = Literal["user"]
-    USER = "user"
-
 
 ChatCompletionUserMessageParamContentTypedDict = TypeAliasType(
     "ChatCompletionUserMessageParamContentTypedDict",

src/openrouter/models/chatstreamcompletioncreateparams.py

@@ -14,7 +14,6 @@ from .responseformatjsonschemaschema import (
     ResponseFormatJSONSchemaSchema,
     ResponseFormatJSONSchemaSchemaTypedDict,
 )
-from enum import Enum
 from openrouter.types import (
     BaseModel,
     Nullable,
@@ -24,17 +23,12 @@ from openrouter.types import (
 )
 import pydantic
 from pydantic import model_serializer
-from typing import Any, Dict, List, Optional, Union
+from typing import Any, Dict, List, Literal, Optional, Union
 from typing_extensions import Annotated, NotRequired, TypeAliasType, TypedDict
 
 
-class ChatStreamCompletionCreateParamsEffort(str, Enum):
+ChatStreamCompletionCreateParamsEffort = Literal["high", "medium", "low", "minimal"]
-    r"""OpenAI-style reasoning effort setting"""
+r"""OpenAI-style reasoning effort setting"""
-
-    HIGH = "high"
-    MEDIUM = "medium"
-    LOW = "low"
-    MINIMAL = "minimal"
 
 
 class ChatStreamCompletionCreateParamsReasoningTypedDict(TypedDict):
@@ -94,8 +88,7 @@ class ChatStreamCompletionCreateParamsReasoning(BaseModel):
         return m
 
 
-class ChatStreamCompletionCreateParamsTypePython(str, Enum):
+ChatStreamCompletionCreateParamsTypePython = Literal["python"]
-    PYTHON = "python"
 
 
 class ChatStreamCompletionCreateParamsResponseFormatPythonTypedDict(TypedDict):
@@ -110,8 +103,7 @@ class ChatStreamCompletionCreateParamsResponseFormatPython(BaseModel):
     type: ChatStreamCompletionCreateParamsTypePython
 
 
-class ChatStreamCompletionCreateParamsTypeGrammar(str, Enum):
+ChatStreamCompletionCreateParamsTypeGrammar = Literal["grammar"]
-    GRAMMAR = "grammar"
 
 
 class ChatStreamCompletionCreateParamsResponseFormatGrammarTypedDict(TypedDict):
@@ -131,8 +123,7 @@ class ChatStreamCompletionCreateParamsResponseFormatGrammar(BaseModel):
     r"""Custom grammar for text generation"""
 
 
-class ChatStreamCompletionCreateParamsTypeJSONSchema(str, Enum):
+ChatStreamCompletionCreateParamsTypeJSONSchema = Literal["json_schema"]
-    JSON_SCHEMA = "json_schema"
 
 
 class ChatStreamCompletionCreateParamsJSONSchemaTypedDict(TypedDict):
@@ -207,8 +198,7 @@ class ChatStreamCompletionCreateParamsResponseFormatJSONSchema(BaseModel):
     json_schema: ChatStreamCompletionCreateParamsJSONSchema
 
 
-class ChatStreamCompletionCreateParamsTypeJSONObject(str, Enum):
+ChatStreamCompletionCreateParamsTypeJSONObject = Literal["json_object"]
-    JSON_OBJECT = "json_object"
 
 
 class ChatStreamCompletionCreateParamsResponseFormatJSONObjectTypedDict(TypedDict):
@@ -223,8 +213,7 @@ class ChatStreamCompletionCreateParamsResponseFormatJSONObject(BaseModel):
     type: ChatStreamCompletionCreateParamsTypeJSONObject
 
 
-class ChatStreamCompletionCreateParamsTypeText(str, Enum):
+ChatStreamCompletionCreateParamsTypeText = Literal["text"]
-    TEXT = "text"
 
 
 class ChatStreamCompletionCreateParamsResponseFormatTextTypedDict(TypedDict):
@@ -291,106 +280,98 @@ class ChatStreamCompletionCreateParamsStreamOptions(BaseModel):
     r"""Include usage information in streaming response"""
 
 
-class ChatStreamCompletionCreateParamsReasoningEffort(str, Enum):
+ChatStreamCompletionCreateParamsReasoningEffort = Literal[
-    r"""Reasoning effort"""
+    "high", "medium", "low", "minimal"
+]
+r"""Reasoning effort"""
 
-    HIGH = "high"
+ChatStreamCompletionCreateParamsDataCollection = Literal["deny", "allow"]
-    MEDIUM = "medium"
+r"""Data collection setting. If no available model provider meets the requirement, your request will return an error.
-    LOW = "low"
+- allow: (default) allow providers which store user data non-transiently and may train on it
-    MINIMAL = "minimal"
+- deny: use only providers which do not collect user data.
 
+"""
 
-class ChatStreamCompletionCreateParamsDataCollection(str, Enum):
+ChatStreamCompletionCreateParamsOrderEnum = Literal[
-    r"""Data collection setting. If no available model provider meets the requirement, your request will return an error.
+    "AnyScale",
-    - allow: (default) allow providers which store user data non-transiently and may train on it
+    "Cent-ML",
-    - deny: use only providers which do not collect user data.
+    "HuggingFace",
-
+    "Hyperbolic 2",
-    """
+    "Lepton",
-
+    "Lynn 2",
-    DENY = "deny"
+    "Lynn",
-    ALLOW = "allow"
+    "Mancer",
-
+    "Modal",
-
+    "OctoAI",
-class ChatStreamCompletionCreateParamsOrderEnum(str, Enum):
+    "Recursal",
-    ANY_SCALE = "AnyScale"
+    "Reflection",
-    CENT_ML = "Cent-ML"
+    "Replicate",
-    HUGGING_FACE = "HuggingFace"
+    "SambaNova 2",
-    HYPERBOLIC_2 = "Hyperbolic 2"
+    "SF Compute",
-    LEPTON = "Lepton"
+    "Together 2",
-    LYNN_2 = "Lynn 2"
+    "01.AI",
-    LYNN = "Lynn"
+    "AI21",
-    MANCER = "Mancer"
+    "AionLabs",
-    MODAL = "Modal"
+    "Alibaba",
-    OCTO_AI = "OctoAI"
+    "Amazon Bedrock",
-    RECURSAL = "Recursal"
+    "Anthropic",
-    REFLECTION = "Reflection"
+    "AtlasCloud",
-    REPLICATE = "Replicate"
+    "Atoma",
-    SAMBA_NOVA_2 = "SambaNova 2"
+    "Avian",
-    SF_COMPUTE = "SF Compute"
+    "Azure",
-    TOGETHER_2 = "Together 2"
+    "BaseTen",
-    ONE_DOT_AI = "01.AI"
+    "Cerebras",
-    AI21 = "AI21"
+    "Chutes",
-    AION_LABS = "AionLabs"
+    "Cloudflare",
-    ALIBABA = "Alibaba"
+    "Cohere",
-    AMAZON_BEDROCK = "Amazon Bedrock"
+    "CrofAI",
-    ANTHROPIC = "Anthropic"
+    "Crusoe",
-    ATLAS_CLOUD = "AtlasCloud"
+    "DeepInfra",
-    ATOMA = "Atoma"
+    "DeepSeek",
-    AVIAN = "Avian"
+    "Enfer",
-    AZURE = "Azure"
+    "Featherless",
-    BASE_TEN = "BaseTen"
+    "Fireworks",
-    CEREBRAS = "Cerebras"
+    "Friendli",
-    CHUTES = "Chutes"
+    "GMICloud",
-    CLOUDFLARE = "Cloudflare"
+    "Google",
-    COHERE = "Cohere"
+    "Google AI Studio",
-    CROF_AI = "CrofAI"
+    "Groq",
-    CRUSOE = "Crusoe"
+    "Hyperbolic",
-    DEEP_INFRA = "DeepInfra"
+    "Inception",
-    DEEP_SEEK = "DeepSeek"
+    "InferenceNet",
-    ENFER = "Enfer"
+    "Infermatic",
-    FEATHERLESS = "Featherless"
+    "Inflection",
-    FIREWORKS = "Fireworks"
+    "InoCloud",
-    FRIENDLI = "Friendli"
+    "Kluster",
-    GMI_CLOUD = "GMICloud"
+    "Lambda",
-    GOOGLE = "Google"
+    "Liquid",
-    GOOGLE_AI_STUDIO = "Google AI Studio"
+    "Mancer 2",
-    GROQ = "Groq"
+    "Meta",
-    HYPERBOLIC = "Hyperbolic"
+    "Minimax",
-    INCEPTION = "Inception"
+    "Mistral",
-    INFERENCE_NET = "InferenceNet"
+    "Moonshot AI",
-    INFERMATIC = "Infermatic"
+    "Morph",
-    INFLECTION = "Inflection"
+    "NCompass",
-    INO_CLOUD = "InoCloud"
+    "Nebius",
-    KLUSTER = "Kluster"
+    "NextBit",
-    LAMBDA = "Lambda"
+    "Nineteen",
-    LIQUID = "Liquid"
+    "Novita",
-    MANCER_2 = "Mancer 2"
+    "OpenAI",
-    META = "Meta"
+    "OpenInference",
-    MINIMAX = "Minimax"
+    "Parasail",
-    MISTRAL = "Mistral"
+    "Perplexity",
-    MOONSHOT_AI = "Moonshot AI"
+    "Phala",
-    MORPH = "Morph"
+    "SambaNova",
-    N_COMPASS = "NCompass"
+    "Stealth",
-    NEBIUS = "Nebius"
+    "Switchpoint",
-    NEXT_BIT = "NextBit"
+    "Targon",
-    NINETEEN = "Nineteen"
+    "Together",
-    NOVITA = "Novita"
+    "Ubicloud",
-    OPEN_AI = "OpenAI"
+    "Venice",
-    OPEN_INFERENCE = "OpenInference"
+    "WandB",
-    PARASAIL = "Parasail"
+    "xAI",
-    PERPLEXITY = "Perplexity"
+    "Z.AI",
-    PHALA = "Phala"
+]
-    SAMBA_NOVA = "SambaNova"
-    STEALTH = "Stealth"
-    SWITCHPOINT = "Switchpoint"
-    TARGON = "Targon"
-    TOGETHER = "Together"
-    UBICLOUD = "Ubicloud"
-    VENICE = "Venice"
-    WAND_B = "WandB"
-    X_AI = "xAI"
-    Z_AI = "Z.AI"
-
 
 ChatStreamCompletionCreateParamsOrderUnionTypedDict = TypeAliasType(
     "ChatStreamCompletionCreateParamsOrderUnionTypedDict",
@@ -404,86 +385,86 @@ ChatStreamCompletionCreateParamsOrderUnion = TypeAliasType(
 )
 
 
-class ChatStreamCompletionCreateParamsOnlyEnum(str, Enum):
+ChatStreamCompletionCreateParamsOnlyEnum = Literal[
-    ANY_SCALE = "AnyScale"
+    "AnyScale",
-    CENT_ML = "Cent-ML"
+    "Cent-ML",
-    HUGGING_FACE = "HuggingFace"
+    "HuggingFace",
-    HYPERBOLIC_2 = "Hyperbolic 2"
+    "Hyperbolic 2",
-    LEPTON = "Lepton"
+    "Lepton",
-    LYNN_2 = "Lynn 2"
+    "Lynn 2",
-    LYNN = "Lynn"
+    "Lynn",
-    MANCER = "Mancer"
+    "Mancer",
-    MODAL = "Modal"
+    "Modal",
-    OCTO_AI = "OctoAI"
+    "OctoAI",
-    RECURSAL = "Recursal"
+    "Recursal",
-    REFLECTION = "Reflection"
+    "Reflection",
-    REPLICATE = "Replicate"
+    "Replicate",
-    SAMBA_NOVA_2 = "SambaNova 2"
+    "SambaNova 2",
-    SF_COMPUTE = "SF Compute"
+    "SF Compute",
-    TOGETHER_2 = "Together 2"
+    "Together 2",
-    ONE_DOT_AI = "01.AI"
+    "01.AI",
-    AI21 = "AI21"
+    "AI21",
-    AION_LABS = "AionLabs"
+    "AionLabs",
-    ALIBABA = "Alibaba"
+    "Alibaba",
-    AMAZON_BEDROCK = "Amazon Bedrock"
+    "Amazon Bedrock",
-    ANTHROPIC = "Anthropic"
+    "Anthropic",
-    ATLAS_CLOUD = "AtlasCloud"
+    "AtlasCloud",
-    ATOMA = "Atoma"
+    "Atoma",
-    AVIAN = "Avian"
+    "Avian",
-    AZURE = "Azure"
+    "Azure",
-    BASE_TEN = "BaseTen"
+    "BaseTen",
-    CEREBRAS = "Cerebras"
+    "Cerebras",
-    CHUTES = "Chutes"
+    "Chutes",
-    CLOUDFLARE = "Cloudflare"
+    "Cloudflare",
-    COHERE = "Cohere"
+    "Cohere",
-    CROF_AI = "CrofAI"
+    "CrofAI",
-    CRUSOE = "Crusoe"
+    "Crusoe",
-    DEEP_INFRA = "DeepInfra"
+    "DeepInfra",
-    DEEP_SEEK = "DeepSeek"
+    "DeepSeek",
-    ENFER = "Enfer"
+    "Enfer",
-    FEATHERLESS = "Featherless"
+    "Featherless",
-    FIREWORKS = "Fireworks"
+    "Fireworks",
-    FRIENDLI = "Friendli"
+    "Friendli",
-    GMI_CLOUD = "GMICloud"
+    "GMICloud",
-    GOOGLE = "Google"
+    "Google",
-    GOOGLE_AI_STUDIO = "Google AI Studio"
+    "Google AI Studio",
-    GROQ = "Groq"
+    "Groq",
-    HYPERBOLIC = "Hyperbolic"
+    "Hyperbolic",
-    INCEPTION = "Inception"
+    "Inception",
-    INFERENCE_NET = "InferenceNet"
+    "InferenceNet",
-    INFERMATIC = "Infermatic"
+    "Infermatic",
-    INFLECTION = "Inflection"
+    "Inflection",
-    INO_CLOUD = "InoCloud"
+    "InoCloud",
-    KLUSTER = "Kluster"
+    "Kluster",
-    LAMBDA = "Lambda"
+    "Lambda",
-    LIQUID = "Liquid"
+    "Liquid",
-    MANCER_2 = "Mancer 2"
+    "Mancer 2",
-    META = "Meta"
+    "Meta",
-    MINIMAX = "Minimax"
+    "Minimax",
-    MISTRAL = "Mistral"
+    "Mistral",
-    MOONSHOT_AI = "Moonshot AI"
+    "Moonshot AI",
-    MORPH = "Morph"
+    "Morph",
-    N_COMPASS = "NCompass"
+    "NCompass",
-    NEBIUS = "Nebius"
+    "Nebius",
-    NEXT_BIT = "NextBit"
+    "NextBit",
-    NINETEEN = "Nineteen"
+    "Nineteen",
-    NOVITA = "Novita"
+    "Novita",
-    OPEN_AI = "OpenAI"
+    "OpenAI",
-    OPEN_INFERENCE = "OpenInference"
+    "OpenInference",
-    PARASAIL = "Parasail"
+    "Parasail",
-    PERPLEXITY = "Perplexity"
+    "Perplexity",
-    PHALA = "Phala"
+    "Phala",
-    SAMBA_NOVA = "SambaNova"
+    "SambaNova",
-    STEALTH = "Stealth"
+    "Stealth",
-    SWITCHPOINT = "Switchpoint"
+    "Switchpoint",
-    TARGON = "Targon"
+    "Targon",
-    TOGETHER = "Together"
+    "Together",
-    UBICLOUD = "Ubicloud"
+    "Ubicloud",
-    VENICE = "Venice"
+    "Venice",
-    WAND_B = "WandB"
+    "WandB",
-    X_AI = "xAI"
+    "xAI",
-    Z_AI = "Z.AI"
+    "Z.AI",
-
+]
 
 ChatStreamCompletionCreateParamsOnlyUnionTypedDict = TypeAliasType(
     "ChatStreamCompletionCreateParamsOnlyUnionTypedDict",
@@ -497,86 +478,86 @@ ChatStreamCompletionCreateParamsOnlyUnion = TypeAliasType(
 )
 
 
-class ChatStreamCompletionCreateParamsIgnoreEnum(str, Enum):
+ChatStreamCompletionCreateParamsIgnoreEnum = Literal[
-    ANY_SCALE = "AnyScale"
+    "AnyScale",
-    CENT_ML = "Cent-ML"
+    "Cent-ML",
-    HUGGING_FACE = "HuggingFace"
+    "HuggingFace",
-    HYPERBOLIC_2 = "Hyperbolic 2"
+    "Hyperbolic 2",
-    LEPTON = "Lepton"
+    "Lepton",
-    LYNN_2 = "Lynn 2"
+    "Lynn 2",
-    LYNN = "Lynn"
+    "Lynn",
-    MANCER = "Mancer"
+    "Mancer",
-    MODAL = "Modal"
+    "Modal",
-    OCTO_AI = "OctoAI"
+    "OctoAI",
-    RECURSAL = "Recursal"
+    "Recursal",
-    REFLECTION = "Reflection"
+    "Reflection",
-    REPLICATE = "Replicate"
+    "Replicate",
-    SAMBA_NOVA_2 = "SambaNova 2"
+    "SambaNova 2",
-    SF_COMPUTE = "SF Compute"
+    "SF Compute",
-    TOGETHER_2 = "Together 2"
+    "Together 2",
-    ONE_DOT_AI = "01.AI"
+    "01.AI",
-    AI21 = "AI21"
+    "AI21",
-    AION_LABS = "AionLabs"
+    "AionLabs",
-    ALIBABA = "Alibaba"
+    "Alibaba",
-    AMAZON_BEDROCK = "Amazon Bedrock"
+    "Amazon Bedrock",
-    ANTHROPIC = "Anthropic"
+    "Anthropic",
-    ATLAS_CLOUD = "AtlasCloud"
+    "AtlasCloud",
-    ATOMA = "Atoma"
+    "Atoma",
-    AVIAN = "Avian"
+    "Avian",
-    AZURE = "Azure"
+    "Azure",
-    BASE_TEN = "BaseTen"
+    "BaseTen",
-    CEREBRAS = "Cerebras"
+    "Cerebras",
-    CHUTES = "Chutes"
+    "Chutes",
-    CLOUDFLARE = "Cloudflare"
+    "Cloudflare",
-    COHERE = "Cohere"
+    "Cohere",
-    CROF_AI = "CrofAI"
+    "CrofAI",
-    CRUSOE = "Crusoe"
+    "Crusoe",
-    DEEP_INFRA = "DeepInfra"
+    "DeepInfra",
-    DEEP_SEEK = "DeepSeek"
+    "DeepSeek",
-    ENFER = "Enfer"
+    "Enfer",
-    FEATHERLESS = "Featherless"
+    "Featherless",
-    FIREWORKS = "Fireworks"
+    "Fireworks",
-    FRIENDLI = "Friendli"
+    "Friendli",
-    GMI_CLOUD = "GMICloud"
+    "GMICloud",
-    GOOGLE = "Google"
+    "Google",
-    GOOGLE_AI_STUDIO = "Google AI Studio"
+    "Google AI Studio",
-    GROQ = "Groq"
+    "Groq",
-    HYPERBOLIC = "Hyperbolic"
+    "Hyperbolic",
-    INCEPTION = "Inception"
+    "Inception",
-    INFERENCE_NET = "InferenceNet"
+    "InferenceNet",
-    INFERMATIC = "Infermatic"
+    "Infermatic",
-    INFLECTION = "Inflection"
+    "Inflection",
-    INO_CLOUD = "InoCloud"
+    "InoCloud",
-    KLUSTER = "Kluster"
+    "Kluster",
-    LAMBDA = "Lambda"
+    "Lambda",
-    LIQUID = "Liquid"
+    "Liquid",
-    MANCER_2 = "Mancer 2"
+    "Mancer 2",
-    META = "Meta"
+    "Meta",
-    MINIMAX = "Minimax"
+    "Minimax",
-    MISTRAL = "Mistral"
+    "Mistral",
-    MOONSHOT_AI = "Moonshot AI"
+    "Moonshot AI",
-    MORPH = "Morph"
+    "Morph",
-    N_COMPASS = "NCompass"
+    "NCompass",
-    NEBIUS = "Nebius"
+    "Nebius",
-    NEXT_BIT = "NextBit"
+    "NextBit",
-    NINETEEN = "Nineteen"
+    "Nineteen",
-    NOVITA = "Novita"
+    "Novita",
-    OPEN_AI = "OpenAI"
+    "OpenAI",
-    OPEN_INFERENCE = "OpenInference"
+    "OpenInference",
-    PARASAIL = "Parasail"
+    "Parasail",
-    PERPLEXITY = "Perplexity"
+    "Perplexity",
-    PHALA = "Phala"
+    "Phala",
-    SAMBA_NOVA = "SambaNova"
+    "SambaNova",
-    STEALTH = "Stealth"
+    "Stealth",
-    SWITCHPOINT = "Switchpoint"
+    "Switchpoint",
-    TARGON = "Targon"
+    "Targon",
-    TOGETHER = "Together"
+    "Together",
-    UBICLOUD = "Ubicloud"
+    "Ubicloud",
-    VENICE = "Venice"
+    "Venice",
-    WAND_B = "WandB"
+    "WandB",
-    X_AI = "xAI"
+    "xAI",
-    Z_AI = "Z.AI"
+    "Z.AI",
-
+]
 
 ChatStreamCompletionCreateParamsIgnoreUnionTypedDict = TypeAliasType(
     "ChatStreamCompletionCreateParamsIgnoreUnionTypedDict",
@@ -590,25 +571,12 @@ ChatStreamCompletionCreateParamsIgnoreUnion = TypeAliasType(
 )
 
 
-class ChatStreamCompletionCreateParamsQuantization(str, Enum):
+ChatStreamCompletionCreateParamsQuantization = Literal[
-    INT4 = "int4"
+    "int4", "int8", "fp4", "fp6", "fp8", "fp16", "bf16", "fp32", "unknown"
-    INT8 = "int8"
+]
-    FP4 = "fp4"
-    FP6 = "fp6"
-    FP8 = "fp8"
-    FP16 = "fp16"
-    BF16 = "bf16"
-    FP32 = "fp32"
-    UNKNOWN = "unknown"
-
-
-class ChatStreamCompletionCreateParamsSort(str, Enum):
-    r"""The sorting strategy to use for this request, if \"order\" is not specified. When set, no load balancing is performed."""
-
-    PRICE = "price"
-    THROUGHPUT = "throughput"
-    LATENCY = "latency"
 
+ChatStreamCompletionCreateParamsSort = Literal["price", "throughput", "latency"]
+r"""The sorting strategy to use for this request, if \"order\" is not specified. When set, no load balancing is performed."""
 
 ChatStreamCompletionCreateParamsPromptTypedDict = TypeAliasType(
     "ChatStreamCompletionCreateParamsPromptTypedDict", Union[float, str, Any]
@@ -817,14 +785,9 @@ class ChatStreamCompletionCreateParamsProvider(BaseModel):
         return m
 
 
-class ChatStreamCompletionCreateParamsIDFileParser(str, Enum):
+ChatStreamCompletionCreateParamsIDFileParser = Literal["file-parser"]
-    FILE_PARSER = "file-parser"
 
-
+ChatStreamCompletionCreateParamsPdfEngine = Literal["mistral-ocr", "pdf-text", "native"]
-class ChatStreamCompletionCreateParamsPdfEngine(str, Enum):
-    MISTRAL_OCR = "mistral-ocr"
-    PDF_TEXT = "pdf-text"
-    NATIVE = "native"
 
 
 class ChatStreamCompletionCreateParamsPdfTypedDict(TypedDict):
@@ -849,8 +812,7 @@ class ChatStreamCompletionCreateParamsPluginFileParser(BaseModel):
     pdf: Optional[ChatStreamCompletionCreateParamsPdf] = None
 
 
-class ChatStreamCompletionCreateParamsIDChainOfThought(str, Enum):
+ChatStreamCompletionCreateParamsIDChainOfThought = Literal["chain-of-thought"]
-    CHAIN_OF_THOUGHT = "chain-of-thought"
 
 
 class ChatStreamCompletionCreateParamsPluginChainOfThoughtTypedDict(TypedDict):
@@ -861,13 +823,9 @@ class ChatStreamCompletionCreateParamsPluginChainOfThought(BaseModel):
     id: ChatStreamCompletionCreateParamsIDChainOfThought
 
 
-class ChatStreamCompletionCreateParamsIDWeb(str, Enum):
+ChatStreamCompletionCreateParamsIDWeb = Literal["web"]
-    WEB = "web"
 
-
+ChatStreamCompletionCreateParamsEngine = Literal["native", "exa"]
-class ChatStreamCompletionCreateParamsEngine(str, Enum):
-    NATIVE = "native"
-    EXA = "exa"
 
 
 class ChatStreamCompletionCreateParamsPluginWebTypedDict(TypedDict):
@@ -887,8 +845,7 @@ class ChatStreamCompletionCreateParamsPluginWeb(BaseModel):
     engine: Optional[ChatStreamCompletionCreateParamsEngine] = None
 
 
-class ChatStreamCompletionCreateParamsIDModeration(str, Enum):
+ChatStreamCompletionCreateParamsIDModeration = Literal["moderation"]
-    MODERATION = "moderation"
 
 
 class ChatStreamCompletionCreateParamsPluginModerationTypedDict(TypedDict):
@@ -969,7 +926,7 @@ class ChatStreamCompletionCreateParamsTypedDict(TypedDict):
     r"""Nucleus sampling parameter (0-1)"""
     user: NotRequired[str]
     r"""Unique user identifier"""
-    models_llm: NotRequired[Nullable[List[str]]]
+    fallback_models: NotRequired[Nullable[List[str]]]
     r"""Order of models to fallback to for this request"""
     reasoning_effort: NotRequired[
         Nullable[ChatStreamCompletionCreateParamsReasoningEffort]
@@ -1050,7 +1007,7 @@ class ChatStreamCompletionCreateParams(BaseModel):
     user: Optional[str] = None
     r"""Unique user identifier"""
 
-    models_llm: Annotated[
+    fallback_models: Annotated[
         OptionalNullable[List[str]], pydantic.Field(alias="models")
     ] = UNSET
     r"""Order of models to fallback to for this request"""
@@ -1089,7 +1046,7 @@ class ChatStreamCompletionCreateParams(BaseModel):
             "tools",
             "top_p",
             "user",
-            "models_llm",
+            "fallback_models",
             "reasoning_effort",
             "provider",
             "plugins",
@@ -1108,7 +1065,7 @@ class ChatStreamCompletionCreateParams(BaseModel):
             "stream_options",
             "temperature",
             "top_p",
-            "models_llm",
+            "fallback_models",
             "reasoning_effort",
             "provider",
         ]

src/openrouter/models/fileannotationdetail.py

@@ -1,18 +1,14 @@
 """Code generated by Speakeasy (https://speakeasy.com). DO NOT EDIT."""
 
 from __future__ import annotations
-from enum import Enum
 from openrouter.types import BaseModel
-from typing import List, Optional, Union
+from typing import List, Literal, Optional, Union
 from typing_extensions import NotRequired, TypeAliasType, TypedDict
 
 
-class TypeFile(str, Enum):
+TypeFile = Literal["file"]
-    FILE = "file"
 
-
+ContentTypeImageURL = Literal["image_url"]
-class ContentTypeImageURL(str, Enum):
-    IMAGE_URL = "image_url"
 
 
 class FileAnnotationDetailImageURLTypedDict(TypedDict):
@@ -34,8 +30,7 @@ class ContentImageURL(BaseModel):
     image_url: FileAnnotationDetailImageURL
 
 
-class ContentTypeText(str, Enum):
+ContentTypeText = Literal["text"]
-    TEXT = "text"
 
 
 class ContentTextTypedDict(TypedDict):

src/openrouter/models/reasoningdetailencrypted.py

@@ -1,7 +1,6 @@
 """Code generated by Speakeasy (https://speakeasy.com). DO NOT EDIT."""
 
 from __future__ import annotations
-from enum import Enum
 from openrouter.types import (
     BaseModel,
     Nullable,
@@ -11,18 +10,15 @@ from openrouter.types import (
 )
 import pydantic
 from pydantic import model_serializer
-from typing import Optional
+from typing import Literal, Optional
 from typing_extensions import Annotated, NotRequired, TypedDict
 
 
-class ReasoningDetailEncryptedType(str, Enum):
+ReasoningDetailEncryptedType = Literal["reasoning.encrypted"]
-    REASONING_ENCRYPTED = "reasoning.encrypted"
 
-
+ReasoningDetailEncryptedFormat = Literal[
-class ReasoningDetailEncryptedFormat(str, Enum):
+    "unknown", "openai-responses-v1", "anthropic-claude-v1"
-    UNKNOWN = "unknown"
+]
-    OPENAI_RESPONSES_V1 = "openai-responses-v1"
-    ANTHROPIC_CLAUDE_V1 = "anthropic-claude-v1"
 
 
 class ReasoningDetailEncryptedTypedDict(TypedDict):
@@ -46,7 +42,7 @@ class ReasoningDetailEncrypted(BaseModel):
 
     format_: Annotated[
         OptionalNullable[ReasoningDetailEncryptedFormat], pydantic.Field(alias="format")
-    ] = ReasoningDetailEncryptedFormat.ANTHROPIC_CLAUDE_V1
+    ] = "anthropic-claude-v1"
 
     index: Optional[float] = None
 

src/openrouter/models/reasoningdetailsummary.py

@@ -1,7 +1,6 @@
 """Code generated by Speakeasy (https://speakeasy.com). DO NOT EDIT."""
 
 from __future__ import annotations
-from enum import Enum
 from openrouter.types import (
     BaseModel,
     Nullable,
@@ -11,18 +10,15 @@ from openrouter.types import (
 )
 import pydantic
 from pydantic import model_serializer
-from typing import Optional
+from typing import Literal, Optional
 from typing_extensions import Annotated, NotRequired, TypedDict
 
 
-class ReasoningDetailSummaryType(str, Enum):
+ReasoningDetailSummaryType = Literal["reasoning.summary"]
-    REASONING_SUMMARY = "reasoning.summary"
 
-
+ReasoningDetailSummaryFormat = Literal[
-class ReasoningDetailSummaryFormat(str, Enum):
+    "unknown", "openai-responses-v1", "anthropic-claude-v1"
-    UNKNOWN = "unknown"
+]
-    OPENAI_RESPONSES_V1 = "openai-responses-v1"
-    ANTHROPIC_CLAUDE_V1 = "anthropic-claude-v1"
 
 
 class ReasoningDetailSummaryTypedDict(TypedDict):
@@ -46,7 +42,7 @@ class ReasoningDetailSummary(BaseModel):
 
     format_: Annotated[
         OptionalNullable[ReasoningDetailSummaryFormat], pydantic.Field(alias="format")
-    ] = ReasoningDetailSummaryFormat.ANTHROPIC_CLAUDE_V1
+    ] = "anthropic-claude-v1"
 
     index: Optional[float] = None
 

src/openrouter/models/reasoningdetailtext.py

@@ -1,7 +1,6 @@
 """Code generated by Speakeasy (https://speakeasy.com). DO NOT EDIT."""
 
 from __future__ import annotations
-from enum import Enum
 from openrouter.types import (
     BaseModel,
     Nullable,
@@ -11,18 +10,15 @@ from openrouter.types import (
 )
 import pydantic
 from pydantic import model_serializer
-from typing import Optional
+from typing import Literal, Optional
 from typing_extensions import Annotated, NotRequired, TypedDict
 
 
-class ReasoningDetailTextType(str, Enum):
+ReasoningDetailTextType = Literal["reasoning.text"]
-    REASONING_TEXT = "reasoning.text"
 
-
+ReasoningDetailTextFormat = Literal[
-class ReasoningDetailTextFormat(str, Enum):
+    "unknown", "openai-responses-v1", "anthropic-claude-v1"
-    UNKNOWN = "unknown"
+]
-    OPENAI_RESPONSES_V1 = "openai-responses-v1"
-    ANTHROPIC_CLAUDE_V1 = "anthropic-claude-v1"
 
 
 class ReasoningDetailTextTypedDict(TypedDict):
@@ -49,7 +45,7 @@ class ReasoningDetailText(BaseModel):
 
     format_: Annotated[
         OptionalNullable[ReasoningDetailTextFormat], pydantic.Field(alias="format")
-    ] = ReasoningDetailTextFormat.ANTHROPIC_CLAUDE_V1
+    ] = "anthropic-claude-v1"
 
     index: Optional[float] = None
 

src/openrouter/models/security.py

@@ -8,11 +8,11 @@ from typing_extensions import Annotated, NotRequired, TypedDict
 
 
 class SecurityTypedDict(TypedDict):
-    bearer_auth: NotRequired[str]
+    api_key: NotRequired[str]
 
 
 class Security(BaseModel):
-    bearer_auth: Annotated[
+    api_key: Annotated[
         Optional[str],
         FieldMetadata(
             security=SecurityMetadata(

src/openrouter/models/urlcitationannotationdetail.py

@@ -1,14 +1,12 @@
 """Code generated by Speakeasy (https://speakeasy.com). DO NOT EDIT."""
 
 from __future__ import annotations
-from enum import Enum
 from openrouter.types import BaseModel
-from typing import Optional
+from typing import Literal, Optional
 from typing_extensions import NotRequired, TypedDict
 
 
-class URLCitationAnnotationDetailType(str, Enum):
+URLCitationAnnotationDetailType = Literal["url_citation"]
-    URL_CITATION = "url_citation"
 
 
 class URLCitationTypedDict(TypedDict):

src/openrouter/sdk.py

@@ -10,6 +10,7 @@ import importlib
 from openrouter import models, utils
 from openrouter._hooks import SDKHooks
 from openrouter.types import OptionalNullable, UNSET
+import sys
 from typing import Any, Callable, Dict, List, Optional, TYPE_CHECKING, Union, cast
 import weakref
 
@@ -30,7 +31,7 @@ class OpenRouter(BaseSDK):
 
     def __init__(
         self,
-        bearer_auth: Optional[Union[Optional[str], Callable[[], Optional[str]]]] = None,
+        api_key: Optional[Union[Optional[str], Callable[[], Optional[str]]]] = None,
         provider_url: Optional[str] = None,
         server_idx: Optional[int] = None,
         server_url: Optional[str] = None,
@@ -43,7 +44,7 @@ class OpenRouter(BaseSDK):
     ) -> None:
         r"""Instantiates the SDK configuring it with the provided parameters.
 
-        :param bearer_auth: The bearer_auth required for authentication
+        :param api_key: The api_key required for authentication
         :param provider_url: Allows setting the provider_url variable for url substitution
         :param server_idx: The index of the server to use for all methods
         :param server_url: The server URL to use for all methods
@@ -75,11 +76,11 @@ class OpenRouter(BaseSDK):
         ), "The provided async_client must implement the AsyncHttpClient protocol."
 
         security: Any = None
-        if callable(bearer_auth):
+        if callable(api_key):
             # pylint: disable=unnecessary-lambda-assignment
-            security = lambda: models.Security(bearer_auth=bearer_auth())
+            security = lambda: models.Security(api_key=api_key())
         else:
-            security = models.Security(bearer_auth=bearer_auth)
+            security = models.Security(api_key=api_key)
 
         if server_url is not None:
             if url_params is not None:
@@ -105,6 +106,7 @@ class OpenRouter(BaseSDK):
                 timeout_ms=timeout_ms,
                 debug_logger=debug_logger,
             ),
+            parent_ref=self,
         )
 
         hooks = SDKHooks()
@@ -124,13 +126,24 @@ class OpenRouter(BaseSDK):
             self.sdk_configuration.async_client_supplied,
         )
 
+    def dynamic_import(self, modname, retries=3):
+        for attempt in range(retries):
+            try:
+                return importlib.import_module(modname)
+            except KeyError:
+                # Clear any half-initialized module and retry
+                sys.modules.pop(modname, None)
+                if attempt == retries - 1:
+                    break
+        raise KeyError(f"Failed to import module '{modname}' after {retries} attempts")
+
     def __getattr__(self, name: str):
         if name in self._sub_sdk_map:
             module_path, class_name = self._sub_sdk_map[name]
             try:
-                module = importlib.import_module(module_path)
+                module = self.dynamic_import(module_path)
                 klass = getattr(module, class_name)
-                instance = klass(self.sdk_configuration)
+                instance = klass(self.sdk_configuration, parent_ref=self)
                 setattr(self, name, instance)
                 return instance
             except ImportError as e:

src/openrouter/utils/__init__.py

@@ -3,6 +3,7 @@
 from typing import TYPE_CHECKING
 from importlib import import_module
 import builtins
+import sys
 
 if TYPE_CHECKING:
     from .annotations import get_discriminator
@@ -162,6 +163,18 @@ _dynamic_imports: dict[str, str] = {
 }
 
 
+def dynamic_import(modname, retries=3):
+    for attempt in range(retries):
+        try:
+            return import_module(modname, __package__)
+        except KeyError:
+            # Clear any half-initialized module and retry
+            sys.modules.pop(modname, None)
+            if attempt == retries - 1:
+                break
+    raise KeyError(f"Failed to import module '{modname}' after {retries} attempts")
+
+
 def __getattr__(attr_name: str) -> object:
     module_name = _dynamic_imports.get(attr_name)
     if module_name is None:
@@ -170,9 +183,8 @@ def __getattr__(attr_name: str) -> object:
         )
 
     try:
-        module = import_module(module_name, __package__)
+        module = dynamic_import(module_name)
-        result = getattr(module, attr_name)
+        return getattr(module, attr_name)
-        return result
     except ImportError as e:
         raise ImportError(
             f"Failed to import {attr_name} from {module_name}: {e}"

src/openrouter/utils/eventstreaming.py

@@ -17,6 +17,9 @@ T = TypeVar("T")
 
 
 class EventStream(Generic[T]):
+    # Holds a reference to the SDK client to avoid it being garbage collected
+    # and cause termination of the underlying httpx client.
+    client_ref: Optional[object]
     response: httpx.Response
     generator: Generator[T, None, None]
 
@@ -25,9 +28,11 @@ class EventStream(Generic[T]):
         response: httpx.Response,
         decoder: Callable[[str], T],
         sentinel: Optional[str] = None,
+        client_ref: Optional[object] = None,
     ):
         self.response = response
         self.generator = stream_events(response, decoder, sentinel)
+        self.client_ref = client_ref
 
     def __iter__(self):
         return self
@@ -43,6 +48,9 @@ class EventStream(Generic[T]):
 
 
 class EventStreamAsync(Generic[T]):
+    # Holds a reference to the SDK client to avoid it being garbage collected
+    # and cause termination of the underlying httpx client.
+    client_ref: Optional[object]
     response: httpx.Response
     generator: AsyncGenerator[T, None]
 
@@ -51,9 +59,11 @@ class EventStreamAsync(Generic[T]):
         response: httpx.Response,
         decoder: Callable[[str], T],
         sentinel: Optional[str] = None,
+        client_ref: Optional[object] = None,
     ):
         self.response = response
         self.generator = stream_events_async(response, decoder, sentinel)
+        self.client_ref = client_ref
 
     def __aiter__(self):
         return self

src/openrouter/utils/security.py

@@ -64,8 +64,8 @@ def get_security_from_env(security: Any, security_class: Any) -> Optional[BaseMo
 
     security_dict: Any = {}
 
-    if os.getenv("OPENROUTER_BEARER_AUTH"):
+    if os.getenv("OPENROUTER_API_KEY"):
-        security_dict["bearer_auth"] = os.getenv("OPENROUTER_BEARER_AUTH")
+        security_dict["api_key"] = os.getenv("OPENROUTER_API_KEY")
 
     return security_class(**security_dict) if security_dict else None
 

uv.lock

@@ -201,7 +201,7 @@ wheels = [
 
 [[package]]
 name = "openrouter"
-version = "0.1.3"
+version = "0.5.2"
 source = { editable = "." }
 dependencies = [
     { name = "httpcore" },