Backport: feat(provider/openai): OpenAILanguageModelOptions type

.changeset/cyan-mirrors-clap.md

@@ -0,0 +1,20 @@
+---
+'@ai-sdk/openai': patch
+---
+
+feat(provider/openai): `OpenAIChatLanguageModelOptions` type
+
+```ts
+import { openai, type OpenAIChatLanguageModelOptions } from '@ai-sdk/openai';
+import { generateText } from 'ai';
+
+await generateText({
+  model: openai.chat('gpt-4o'),
+  prompt: 'Invent a new holiday and describe its traditions.',
+  providerOptions: {
+    openai: {
+      user: 'user-123',
+    } satisfies OpenAIChatLanguageModelOptions,
+  },
+});
+```

content/providers/01-ai-sdk-providers/03-openai.mdx

@@ -624,6 +624,8 @@ OpenAI chat models support also some model specific provider options that are no
 You can pass them in the `providerOptions` argument:
 
 ```ts
+import { openai, type OpenAIChatLanguageModelOptions } from '@ai-sdk/openai';
+
 const model = openai.chat('gpt-5');
 
 await generateText({
@@ -635,7 +637,7 @@ await generateText({
         '50256': -100,
       },
       user: 'test-user', // optional unique user identifier
-    },
+    } satisfies OpenAIChatLanguageModelOptions,
   },
 });
 ```

examples/ai-core/src/generate-text/openai-provider-options.ts

@@ -0,0 +1,33 @@
+import { openai, type OpenAIChatLanguageModelOptions } from '@ai-sdk/openai';
+import { generateText } from 'ai';
+import 'dotenv/config';
+
+async function main() {
+  const { text, usage } = await generateText({
+    model: openai.chat('gpt-4o'),
+    prompt: 'Invent a new holiday and describe its traditions.',
+    providerOptions: {
+      openai: {
+        logitBias: {},
+        logprobs: 1,
+        user: '<user_id>',
+        maxCompletionTokens: 100,
+        store: false,
+        structuredOutputs: false,
+        serviceTier: 'auto',
+        strictJsonSchema: false,
+        textVerbosity: 'medium',
+        promptCacheKey: '<prompt_cache_key>',
+        safetyIdentifier: '<safety_identifier>',
+        // @ts-expect-error
+        invalidOption: null,
+      } satisfies OpenAIChatLanguageModelOptions,
+    },
+  });
+
+  console.log(text);
+  console.log();
+  console.log('Usage:', usage);
+}
+
+main().catch(console.error);

packages/openai/src/chat/openai-chat-language-model.ts

@@ -30,7 +30,7 @@ import { getResponseMetadata } from './get-response-metadata';
 import { mapOpenAIFinishReason } from './map-openai-finish-reason';
 import {
   OpenAIChatModelId,
-  openaiProviderOptions,
+  openaiChatLanguageModelOptions,
 } from './openai-chat-options';
 import { prepareChatTools } from './openai-chat-prepare-tools';
 
@@ -83,7 +83,7 @@ export class OpenAIChatLanguageModel implements LanguageModelV2 {
       (await parseProviderOptions({
         provider: 'openai',
         providerOptions,
-        schema: openaiProviderOptions,
+        schema: openaiChatLanguageModelOptions,
       })) ?? {};
 
     const structuredOutputs = openaiOptions.structuredOutputs ?? true;

packages/openai/src/chat/openai-chat-options.ts

@@ -39,7 +39,7 @@ export type OpenAIChatModelId =
   | 'gpt-5-chat-latest'
   | (string & {});
 
-export const openaiProviderOptions = z.object({
+export const openaiChatLanguageModelOptions = z.object({
   /**
    * Modify the likelihood of specified tokens appearing in the completion.
    *
@@ -141,4 +141,6 @@ export const openaiProviderOptions = z.object({
   safetyIdentifier: z.string().optional(),
 });
 
-export type OpenAIProviderOptions = z.infer<typeof openaiProviderOptions>;
+export type OpenAIChatLanguageModelOptions = z.infer<
+  typeof openaiChatLanguageModelOptions
+>;

packages/openai/src/index.ts

@@ -1,4 +1,5 @@
 export { createOpenAI, openai } from './openai-provider';
 export type { OpenAIProvider, OpenAIProviderSettings } from './openai-provider';
 export type { OpenAIResponsesProviderOptions } from './responses/openai-responses-language-model';
+export type { OpenAIChatLanguageModelOptions } from './chat/openai-chat-options';
 export { VERSION } from './version';