@ai-sdk/openai 1.0.7 → 1.0.9

package/CHANGELOG.md

@@ -1,5 +1,22 @@
 # @ai-sdk/openai
 
+## 1.0.9
+
+### Patch Changes
+
+- 3fab0fb: feat (provider/openai): support reasoning_effort setting
+- e956eed: feat (provider/openai): update model list and add o1
+- 6faab13: feat (provider/openai): simulated streaming setting
+
+## 1.0.8
+
+### Patch Changes
+
+- 09a9cab: feat (ai/core): add experimental generateImage function
+- Updated dependencies [09a9cab]
+  - @ai-sdk/provider@1.0.2
+  - @ai-sdk/provider-utils@2.0.4
+
 ## 1.0.7
 
 ### Patch Changes

package/dist/index.d.mts

@@ -1,7 +1,7 @@
-import { LanguageModelV1, ProviderV1, EmbeddingModelV1 } from '@ai-sdk/provider';
+import { LanguageModelV1, ProviderV1, EmbeddingModelV1, ImageModelV1 } from '@ai-sdk/provider';
 import { FetchFunction } from '@ai-sdk/provider-utils';
 
-type OpenAIChatModelId = 'o1-preview' | 'o1-mini' | 'gpt-4o' | 'gpt-4o-2024-05-13' | 'gpt-4o-2024-08-06' | 'gpt-4o-2024-11-20' | 'gpt-4o-audio-preview' | 'gpt-4o-audio-preview-2024-10-01' | 'gpt-4o-mini' | 'gpt-4o-mini-2024-07-18' | 'gpt-4-turbo' | 'gpt-4-turbo-2024-04-09' | 'gpt-4-turbo-preview' | 'gpt-4-0125-preview' | 'gpt-4-1106-preview' | 'gpt-4' | 'gpt-4-0613' | 'gpt-3.5-turbo-0125' | 'gpt-3.5-turbo' | 'gpt-3.5-turbo-1106' | (string & {});
+type OpenAIChatModelId = 'o1' | 'o1-2024-12-17' | 'o1-mini' | 'o1-mini-2024-09-12' | 'o1-preview' | 'o1-preview-2024-09-12' | 'gpt-4o' | 'gpt-4o-2024-05-13' | 'gpt-4o-2024-08-06' | 'gpt-4o-2024-11-20' | 'gpt-4o-audio-preview' | 'gpt-4o-audio-preview-2024-10-01' | 'gpt-4o-audio-preview-2024-12-17' | 'gpt-4o-mini' | 'gpt-4o-mini-2024-07-18' | 'gpt-4-turbo' | 'gpt-4-turbo-2024-04-09' | 'gpt-4-turbo-preview' | 'gpt-4-0125-preview' | 'gpt-4-1106-preview' | 'gpt-4' | 'gpt-4-0613' | 'gpt-3.5-turbo-0125' | 'gpt-3.5-turbo' | 'gpt-3.5-turbo-1106' | (string & {});
 interface OpenAIChatSettings {
     /**
   Modify the likelihood of specified tokens appearing in the completion.
@@ -65,6 +65,17 @@ interface OpenAIChatSettings {
   Defaults to `false`.
      */
     downloadImages?: boolean;
+    /**
+  Simulates streaming by using a normal generate call and returning it as a stream.
+  Enable this if the model that you are using does not support streaming.
+  
+  Defaults to `false`.
+     */
+    simulateStreaming?: boolean;
+    /**
+  Reasoning effort for reasoning models. Defaults to `medium`.
+     */
+    reasoningEffort?: 'low' | 'medium' | 'high';
 }
 
 type OpenAICompletionModelId = 'gpt-3.5-turbo-instruct' | (string & {});
@@ -156,6 +167,8 @@ interface OpenAIEmbeddingSettings {
     user?: string;
 }
 
+type OpenAIImageModelId = 'dall-e-3' | 'dall-e-2' | (string & {});
+
 interface OpenAIProvider extends ProviderV1 {
     (modelId: 'gpt-3.5-turbo-instruct', settings?: OpenAICompletionSettings): OpenAICompletionLanguageModel;
     (modelId: OpenAIChatModelId, settings?: OpenAIChatSettings): LanguageModelV1;
@@ -186,6 +199,10 @@ interface OpenAIProvider extends ProviderV1 {
   Creates a model for text embeddings.
      */
     textEmbeddingModel(modelId: OpenAIEmbeddingModelId, settings?: OpenAIEmbeddingSettings): EmbeddingModelV1<string>;
+    /**
+  Creates a model for image generation.
+     */
+    image(modelId: OpenAIImageModelId): ImageModelV1;
 }
 interface OpenAIProviderSettings {
     /**

package/dist/index.d.ts

@@ -1,7 +1,7 @@
-import { LanguageModelV1, ProviderV1, EmbeddingModelV1 } from '@ai-sdk/provider';
+import { LanguageModelV1, ProviderV1, EmbeddingModelV1, ImageModelV1 } from '@ai-sdk/provider';
 import { FetchFunction } from '@ai-sdk/provider-utils';
 
-type OpenAIChatModelId = 'o1-preview' | 'o1-mini' | 'gpt-4o' | 'gpt-4o-2024-05-13' | 'gpt-4o-2024-08-06' | 'gpt-4o-2024-11-20' | 'gpt-4o-audio-preview' | 'gpt-4o-audio-preview-2024-10-01' | 'gpt-4o-mini' | 'gpt-4o-mini-2024-07-18' | 'gpt-4-turbo' | 'gpt-4-turbo-2024-04-09' | 'gpt-4-turbo-preview' | 'gpt-4-0125-preview' | 'gpt-4-1106-preview' | 'gpt-4' | 'gpt-4-0613' | 'gpt-3.5-turbo-0125' | 'gpt-3.5-turbo' | 'gpt-3.5-turbo-1106' | (string & {});
+type OpenAIChatModelId = 'o1' | 'o1-2024-12-17' | 'o1-mini' | 'o1-mini-2024-09-12' | 'o1-preview' | 'o1-preview-2024-09-12' | 'gpt-4o' | 'gpt-4o-2024-05-13' | 'gpt-4o-2024-08-06' | 'gpt-4o-2024-11-20' | 'gpt-4o-audio-preview' | 'gpt-4o-audio-preview-2024-10-01' | 'gpt-4o-audio-preview-2024-12-17' | 'gpt-4o-mini' | 'gpt-4o-mini-2024-07-18' | 'gpt-4-turbo' | 'gpt-4-turbo-2024-04-09' | 'gpt-4-turbo-preview' | 'gpt-4-0125-preview' | 'gpt-4-1106-preview' | 'gpt-4' | 'gpt-4-0613' | 'gpt-3.5-turbo-0125' | 'gpt-3.5-turbo' | 'gpt-3.5-turbo-1106' | (string & {});
 interface OpenAIChatSettings {
     /**
   Modify the likelihood of specified tokens appearing in the completion.
@@ -65,6 +65,17 @@ interface OpenAIChatSettings {
   Defaults to `false`.
      */
     downloadImages?: boolean;
+    /**
+  Simulates streaming by using a normal generate call and returning it as a stream.
+  Enable this if the model that you are using does not support streaming.
+  
+  Defaults to `false`.
+     */
+    simulateStreaming?: boolean;
+    /**
+  Reasoning effort for reasoning models. Defaults to `medium`.
+     */
+    reasoningEffort?: 'low' | 'medium' | 'high';
 }
 
 type OpenAICompletionModelId = 'gpt-3.5-turbo-instruct' | (string & {});
@@ -156,6 +167,8 @@ interface OpenAIEmbeddingSettings {
     user?: string;
 }
 
+type OpenAIImageModelId = 'dall-e-3' | 'dall-e-2' | (string & {});
+
 interface OpenAIProvider extends ProviderV1 {
     (modelId: 'gpt-3.5-turbo-instruct', settings?: OpenAICompletionSettings): OpenAICompletionLanguageModel;
     (modelId: OpenAIChatModelId, settings?: OpenAIChatSettings): LanguageModelV1;
@@ -186,6 +199,10 @@ interface OpenAIProvider extends ProviderV1 {
   Creates a model for text embeddings.
      */
     textEmbeddingModel(modelId: OpenAIEmbeddingModelId, settings?: OpenAIEmbeddingSettings): EmbeddingModelV1<string>;
+    /**
+  Creates a model for image generation.
+     */
+    image(modelId: OpenAIImageModelId): ImageModelV1;
 }
 interface OpenAIProviderSettings {
     /**

package/dist/index.js

@@ -26,7 +26,7 @@ __export(src_exports, {
 module.exports = __toCommonJS(src_exports);
 
 // src/openai-provider.ts
-var import_provider_utils6 = require("@ai-sdk/provider-utils");
+var import_provider_utils7 = require("@ai-sdk/provider-utils");
 
 // src/openai-chat-language-model.ts
 var import_provider3 = require("@ai-sdk/provider");
@@ -376,7 +376,7 @@ var OpenAIChatLanguageModel = class {
     seed,
     providerMetadata
   }) {
-    var _a, _b, _c, _d, _e, _f, _g, _h, _i, _j;
+    var _a, _b, _c, _d, _e, _f, _g, _h;
     const type = mode.type;
     const warnings = [];
     if (topK != null) {
@@ -430,10 +430,11 @@ var OpenAIChatLanguageModel = class {
       stop: stopSequences,
       seed,
       // openai specific settings:
-      max_completion_tokens: (_c = (_b = providerMetadata == null ? void 0 : providerMetadata.openai) == null ? void 0 : _b.maxCompletionTokens) != null ? _c : void 0,
-      store: (_e = (_d = providerMetadata == null ? void 0 : providerMetadata.openai) == null ? void 0 : _d.store) != null ? _e : void 0,
-      metadata: (_g = (_f = providerMetadata == null ? void 0 : providerMetadata.openai) == null ? void 0 : _f.metadata) != null ? _g : void 0,
-      prediction: (_i = (_h = providerMetadata == null ? void 0 : providerMetadata.openai) == null ? void 0 : _h.prediction) != null ? _i : void 0,
+      max_completion_tokens: (_b = providerMetadata == null ? void 0 : providerMetadata.openai) == null ? void 0 : _b.maxCompletionTokens,
+      store: (_c = providerMetadata == null ? void 0 : providerMetadata.openai) == null ? void 0 : _c.store,
+      metadata: (_d = providerMetadata == null ? void 0 : providerMetadata.openai) == null ? void 0 : _d.metadata,
+      prediction: (_e = providerMetadata == null ? void 0 : providerMetadata.openai) == null ? void 0 : _e.prediction,
+      reasoning_effort: (_g = (_f = providerMetadata == null ? void 0 : providerMetadata.openai) == null ? void 0 : _f.reasoningEffort) != null ? _g : this.settings.reasoningEffort,
       // messages:
       messages: convertToOpenAIChatMessages({
         prompt,
@@ -473,7 +474,7 @@ var OpenAIChatLanguageModel = class {
               json_schema: {
                 schema: mode.schema,
                 strict: true,
-                name: (_j = mode.name) != null ? _j : "response",
+                name: (_h = mode.name) != null ? _h : "response",
                 description: mode.description
               }
             } : { type: "json_object" }
@@ -584,6 +585,42 @@ var OpenAIChatLanguageModel = class {
     };
   }
   async doStream(options) {
+    if (this.settings.simulateStreaming) {
+      const result = await this.doGenerate(options);
+      const simulatedStream = new ReadableStream({
+        start(controller) {
+          controller.enqueue({ type: "response-metadata", ...result.response });
+          if (result.text) {
+            controller.enqueue({
+              type: "text-delta",
+              textDelta: result.text
+            });
+          }
+          if (result.toolCalls) {
+            for (const toolCall of result.toolCalls) {
+              controller.enqueue({
+                type: "tool-call",
+                ...toolCall
+              });
+            }
+          }
+          controller.enqueue({
+            type: "finish",
+            finishReason: result.finishReason,
+            usage: result.usage,
+            logprobs: result.logprobs,
+            providerMetadata: result.providerMetadata
+          });
+          controller.close();
+        }
+      });
+      return {
+        stream: simulatedStream,
+        rawCall: result.rawCall,
+        rawResponse: result.rawResponse,
+        warnings: result.warnings
+      };
+    }
     const { args, warnings } = this.getArgs(options);
     const body = {
       ...args,
@@ -1336,14 +1373,65 @@ var openaiTextEmbeddingResponseSchema = import_zod4.z.object({
   usage: import_zod4.z.object({ prompt_tokens: import_zod4.z.number() }).nullish()
 });
 
+// src/openai-image-model.ts
+var import_provider_utils6 = require("@ai-sdk/provider-utils");
+var import_zod5 = require("zod");
+var OpenAIImageModel = class {
+  constructor(modelId, config) {
+    this.specificationVersion = "v1";
+    this.modelId = modelId;
+    this.config = config;
+  }
+  get provider() {
+    return this.config.provider;
+  }
+  async doGenerate({
+    prompt,
+    n,
+    size,
+    providerOptions,
+    headers,
+    abortSignal
+  }) {
+    var _a;
+    const { value: response } = await (0, import_provider_utils6.postJsonToApi)({
+      url: this.config.url({
+        path: "/images/generations",
+        modelId: this.modelId
+      }),
+      headers: (0, import_provider_utils6.combineHeaders)(this.config.headers(), headers),
+      body: {
+        model: this.modelId,
+        prompt,
+        n,
+        size,
+        ...(_a = providerOptions.openai) != null ? _a : {},
+        response_format: "b64_json"
+      },
+      failedResponseHandler: openaiFailedResponseHandler,
+      successfulResponseHandler: (0, import_provider_utils6.createJsonResponseHandler)(
+        openaiImageResponseSchema
+      ),
+      abortSignal,
+      fetch: this.config.fetch
+    });
+    return {
+      images: response.data.map((item) => item.b64_json)
+    };
+  }
+};
+var openaiImageResponseSchema = import_zod5.z.object({
+  data: import_zod5.z.array(import_zod5.z.object({ b64_json: import_zod5.z.string() }))
+});
+
 // src/openai-provider.ts
 function createOpenAI(options = {}) {
   var _a, _b, _c;
-  const baseURL = (_a = (0, import_provider_utils6.withoutTrailingSlash)(options.baseURL)) != null ? _a : "https://api.openai.com/v1";
+  const baseURL = (_a = (0, import_provider_utils7.withoutTrailingSlash)(options.baseURL)) != null ? _a : "https://api.openai.com/v1";
   const compatibility = (_b = options.compatibility) != null ? _b : "compatible";
   const providerName = (_c = options.name) != null ? _c : "openai";
   const getHeaders = () => ({
-    Authorization: `Bearer ${(0, import_provider_utils6.loadApiKey)({
+    Authorization: `Bearer ${(0, import_provider_utils7.loadApiKey)({
       apiKey: options.apiKey,
       environmentVariableName: "OPENAI_API_KEY",
       description: "OpenAI"
@@ -1372,6 +1460,12 @@ function createOpenAI(options = {}) {
     headers: getHeaders,
     fetch: options.fetch
   });
+  const createImageModel = (modelId) => new OpenAIImageModel(modelId, {
+    provider: `${providerName}.image`,
+    url: ({ path }) => `${baseURL}${path}`,
+    headers: getHeaders,
+    fetch: options.fetch
+  });
   const createLanguageModel = (modelId, settings) => {
     if (new.target) {
       throw new Error(
@@ -1395,6 +1489,7 @@ function createOpenAI(options = {}) {
   provider.embedding = createEmbeddingModel;
   provider.textEmbedding = createEmbeddingModel;
   provider.textEmbeddingModel = createEmbeddingModel;
+  provider.image = createImageModel;
   return provider;
 }
 var openai = createOpenAI({