@ai-sdk/openai 2.1.0-beta.11 → 2.1.0-beta.12

package/CHANGELOG.md

@@ -1,5 +1,17 @@
 # @ai-sdk/openai
 
+## 2.1.0-beta.12
+
+### Patch Changes
+
+- 046aa3b: feat(provider): speech model v3 spec
+- e8109d3: feat: tool execution approval
+- 21e20c0: feat(provider): transcription model v3 spec
+- Updated dependencies [046aa3b]
+- Updated dependencies [e8109d3]
+  - @ai-sdk/provider@2.1.0-beta.5
+  - @ai-sdk/provider-utils@3.1.0-beta.7
+
 ## 2.1.0-beta.11
 
 ### Patch Changes

package/dist/index.d.mts

@@ -1,4 +1,4 @@
-import { ProviderV3, LanguageModelV3, EmbeddingModelV3, ImageModelV3, TranscriptionModelV2, SpeechModelV2 } from '@ai-sdk/provider';
+import { ProviderV3, LanguageModelV3, EmbeddingModelV3, ImageModelV3, TranscriptionModelV3, SpeechModelV3 } from '@ai-sdk/provider';
 import * as _ai_sdk_provider_utils from '@ai-sdk/provider-utils';
 import { FetchFunction } from '@ai-sdk/provider-utils';
 import { z } from 'zod/v4';
@@ -311,11 +311,11 @@ interface OpenAIProvider extends ProviderV3 {
     /**
   Creates a model for transcription.
      */
-    transcription(modelId: OpenAITranscriptionModelId): TranscriptionModelV2;
+    transcription(modelId: OpenAITranscriptionModelId): TranscriptionModelV3;
     /**
   Creates a model for speech generation.
      */
-    speech(modelId: OpenAISpeechModelId): SpeechModelV2;
+    speech(modelId: OpenAISpeechModelId): SpeechModelV3;
     /**
   OpenAI-specific tools.
      */

package/dist/index.d.ts

@@ -1,4 +1,4 @@
-import { ProviderV3, LanguageModelV3, EmbeddingModelV3, ImageModelV3, TranscriptionModelV2, SpeechModelV2 } from '@ai-sdk/provider';
+import { ProviderV3, LanguageModelV3, EmbeddingModelV3, ImageModelV3, TranscriptionModelV3, SpeechModelV3 } from '@ai-sdk/provider';
 import * as _ai_sdk_provider_utils from '@ai-sdk/provider-utils';
 import { FetchFunction } from '@ai-sdk/provider-utils';
 import { z } from 'zod/v4';
@@ -311,11 +311,11 @@ interface OpenAIProvider extends ProviderV3 {
     /**
   Creates a model for transcription.
      */
-    transcription(modelId: OpenAITranscriptionModelId): TranscriptionModelV2;
+    transcription(modelId: OpenAITranscriptionModelId): TranscriptionModelV3;
     /**
   Creates a model for speech generation.
      */
-    speech(modelId: OpenAISpeechModelId): SpeechModelV2;
+    speech(modelId: OpenAISpeechModelId): SpeechModelV3;
     /**
   OpenAI-specific tools.
      */

package/dist/index.js

@@ -60,6 +60,7 @@ function convertToOpenAIChatMessages({
   prompt,
   systemMessageMode = "system"
 }) {
+  var _a;
   const messages = [];
   const warnings = [];
   for (const { role, content } of prompt) {
@@ -98,7 +99,7 @@ function convertToOpenAIChatMessages({
         messages.push({
           role: "user",
           content: content.map((part, index) => {
-            var _a, _b, _c;
+            var _a2, _b, _c;
             switch (part.type) {
               case "text": {
                 return { type: "text", text: part.text };
@@ -111,7 +112,7 @@ function convertToOpenAIChatMessages({
                     image_url: {
                       url: part.data instanceof URL ? part.data.toString() : `data:${mediaType};base64,${(0, import_provider_utils2.convertToBase64)(part.data)}`,
                       // OpenAI specific extension: image detail
-                      detail: (_b = (_a = part.providerOptions) == null ? void 0 : _a.openai) == null ? void 0 : _b.imageDetail
+                      detail: (_b = (_a2 = part.providerOptions) == null ? void 0 : _a2.openai) == null ? void 0 : _b.imageDetail
                     }
                   };
                 } else if (part.mediaType.startsWith("audio/")) {
@@ -208,6 +209,9 @@ function convertToOpenAIChatMessages({
             case "error-text":
               contentValue = output.value;
               break;
+            case "execution-denied":
+              contentValue = (_a = output.reason) != null ? _a : "Tool execution denied.";
+              break;
             case "content":
             case "json":
             case "error-json":
@@ -2005,7 +2009,7 @@ async function convertToOpenAIResponsesInput({
   store,
   hasLocalShellTool = false
 }) {
-  var _a, _b, _c, _d, _e, _f, _g, _h, _i;
+  var _a, _b, _c, _d, _e, _f, _g, _h, _i, _j;
   const input = [];
   const warnings = [];
   for (const { role, content } of prompt) {
@@ -2208,6 +2212,9 @@ async function convertToOpenAIResponsesInput({
             case "error-text":
               contentValue = output.value;
               break;
+            case "execution-denied":
+              contentValue = (_j = output.reason) != null ? _j : "Tool execution denied.";
+              break;
             case "content":
             case "json":
             case "error-json":
@@ -3781,7 +3788,7 @@ var OpenAISpeechModel = class {
   constructor(modelId, config) {
     this.modelId = modelId;
     this.config = config;
-    this.specificationVersion = "v2";
+    this.specificationVersion = "v3";
   }
   get provider() {
     return this.config.provider;
@@ -3972,7 +3979,7 @@ var OpenAITranscriptionModel = class {
   constructor(modelId, config) {
     this.modelId = modelId;
     this.config = config;
-    this.specificationVersion = "v2";
+    this.specificationVersion = "v3";
   }
   get provider() {
     return this.config.provider;
@@ -4102,7 +4109,7 @@ var openaiTranscriptionResponseSchema = import_v419.z.object({
 });
 
 // src/version.ts
-var VERSION = true ? "2.1.0-beta.11" : "0.0.0-test";
+var VERSION = true ? "2.1.0-beta.12" : "0.0.0-test";
 
 // src/openai-provider.ts
 function createOpenAI(options = {}) {