@ai-sdk/xai 4.0.0-canary.70 → 4.0.0-canary.72

package/CHANGELOG.md

@@ -1,5 +1,33 @@
 # @ai-sdk/xai
 
+## 4.0.0-canary.72
+
+### Patch Changes
+
+- 7486744: Add xAI speech-to-text transcription support.
+- 7486744: feat(provider/xai): add text-to-speech support
+
+## 4.0.0-canary.71
+
+### Patch Changes
+
+- ce769dd: feat(provider): add experimental Realtime API support for voice conversations
+
+  Adds first-class support for realtime (speech-to-speech) APIs:
+
+  - `Experimental_RealtimeModelV4` spec in `@ai-sdk/provider` with normalized event types and factory
+  - OpenAI, Google, and xAI realtime provider implementations
+  - `openai.experimental_realtime()` / `google.experimental_realtime()` / `xai.experimental_realtime()` work in both server and browser
+  - `.getToken()` static method on each provider for server-side ephemeral token creation
+  - `experimental_getRealtimeToolDefinitions` helper for provider session tool definitions
+  - `experimental_useRealtime` hook in `@ai-sdk/react` returning `UIMessage[]` (aligned with `useChat`), with `onToolCall` and `addToolOutput` for client-driven tool execution
+  - `inputAudioTranscription` session config for showing transcribed user audio messages when supported by the provider
+
+- Updated dependencies [ce769dd]
+  - @ai-sdk/provider@4.0.0-canary.18
+  - @ai-sdk/openai-compatible@3.0.0-canary.54
+  - @ai-sdk/provider-utils@5.0.0-canary.46
+
 ## 4.0.0-canary.70
 
 ### Patch Changes

package/dist/index.d.ts

@@ -1,7 +1,7 @@
 import { z } from 'zod/v4';
 import * as _ai_sdk_provider_utils from '@ai-sdk/provider-utils';
 import { InferSchema, FetchFunction } from '@ai-sdk/provider-utils';
-import { ProviderV4, LanguageModelV4, ImageModelV4, Experimental_VideoModelV4, FilesV4 } from '@ai-sdk/provider';
+import { ProviderV4, LanguageModelV4, ImageModelV4, Experimental_VideoModelV4, Experimental_RealtimeFactoryV4, SpeechModelV4, TranscriptionModelV4, FilesV4, Experimental_RealtimeModelV4, Experimental_RealtimeModelV4ClientSecretOptions, Experimental_RealtimeModelV4ClientSecretResult, Experimental_RealtimeModelV4ServerEvent, Experimental_RealtimeModelV4ClientEvent, Experimental_RealtimeModelV4SessionConfig } from '@ai-sdk/provider';
 
 type XaiChatModelId = 'grok-4.20-non-reasoning' | 'grok-4.20-reasoning' | 'grok-4.3' | 'grok-latest' | (string & {});
 declare const xaiLanguageModelChatOptions: z.ZodObject<{
@@ -179,6 +179,27 @@ interface XaiLegacyReferenceToVideoOptions extends XaiVideoSharedOptions {
  */
 type XaiVideoModelOptions = XaiVideoGenerationOptions | XaiVideoEditModeOptions | XaiVideoExtendModeOptions | XaiVideoReferenceToVideoOptions | XaiLegacyEditVideoOptions | XaiLegacyReferenceToVideoOptions;
 
+declare const xaiSpeechModelOptionsSchema: _ai_sdk_provider_utils.LazySchema<{
+    sampleRate?: 8000 | 16000 | 22050 | 24000 | 44100 | 48000 | null | undefined;
+    bitRate?: 32000 | 64000 | 96000 | 128000 | 192000 | null | undefined;
+    optimizeStreamingLatency?: 0 | 1 | 2 | null | undefined;
+    textNormalization?: boolean | null | undefined;
+}>;
+type XaiSpeechModelOptions = InferSchema<typeof xaiSpeechModelOptionsSchema>;
+
+declare const xaiTranscriptionModelOptionsSchema: _ai_sdk_provider_utils.LazySchema<{
+    audioFormat?: "pcm" | "mulaw" | "alaw" | null | undefined;
+    sampleRate?: 8000 | 16000 | 22050 | 24000 | 44100 | 48000 | null | undefined;
+    language?: string | null | undefined;
+    format?: boolean | null | undefined;
+    multichannel?: boolean | null | undefined;
+    channels?: number | null | undefined;
+    diarize?: boolean | null | undefined;
+    keyterm?: string | string[] | null | undefined;
+    fillerWords?: boolean | null | undefined;
+}>;
+type XaiTranscriptionModelOptions = InferSchema<typeof xaiTranscriptionModelOptionsSchema>;
+
 declare const xaiFilesOptionsSchema: _ai_sdk_provider_utils.LazySchema<{
     [x: string]: unknown;
     teamId?: string | undefined;
@@ -419,6 +440,23 @@ interface XaiProvider extends ProviderV4 {
      * Creates an Xai video model for video generation.
      */
     videoModel(modelId: XaiVideoModelId): Experimental_VideoModelV4;
+    experimental_realtime: Experimental_RealtimeFactoryV4;
+    /**
+     * Creates an xAI model for speech generation (text-to-speech).
+     */
+    speech(): SpeechModelV4;
+    /**
+     * Creates an xAI model for speech generation (text-to-speech).
+     */
+    speechModel(): SpeechModelV4;
+    /**
+     * Creates an xAI model for speech-to-text transcription.
+     */
+    transcription(): TranscriptionModelV4;
+    /**
+     * Creates an xAI model for speech-to-text transcription.
+     */
+    transcriptionModel(): TranscriptionModelV4;
     /**
      * Returns the xAI files interface for uploading files.
      */
@@ -454,6 +492,31 @@ interface XaiProviderSettings {
 declare function createXai(options?: XaiProviderSettings): XaiProvider;
 declare const xai: XaiProvider;
 
+type XaiRealtimeModelConfig = {
+    provider: string;
+    baseURL: string;
+    headers: () => Record<string, string | undefined>;
+    fetch?: FetchFunction;
+};
+declare class XaiRealtimeModel implements Experimental_RealtimeModelV4 {
+    readonly specificationVersion: "v4";
+    readonly provider: string;
+    readonly modelId: string;
+    private readonly config;
+    constructor(modelId: string, config: XaiRealtimeModelConfig);
+    doCreateClientSecret(options: Experimental_RealtimeModelV4ClientSecretOptions): Promise<Experimental_RealtimeModelV4ClientSecretResult>;
+    getWebSocketConfig(options: {
+        token: string;
+        url: string;
+    }): {
+        url: string;
+        protocols?: string[];
+    };
+    parseServerEvent(raw: unknown): Experimental_RealtimeModelV4ServerEvent;
+    serializeClientEvent(event: Experimental_RealtimeModelV4ClientEvent): unknown;
+    buildSessionConfig(config: Experimental_RealtimeModelV4SessionConfig): Record<string, unknown>;
+}
+
 declare const VERSION: string;
 
-export { VERSION, type XaiErrorData, type XaiFilesOptions, type XaiImageModelOptions, type XaiImageModelOptions as XaiImageProviderOptions, type XaiLanguageModelChatOptions, type XaiLanguageModelResponsesOptions, type XaiProvider, type XaiLanguageModelChatOptions as XaiProviderOptions, type XaiProviderSettings, type XaiLanguageModelResponsesOptions as XaiResponsesProviderOptions, type XaiVideoModelId, type XaiVideoModelOptions, type XaiVideoModelOptions as XaiVideoProviderOptions, codeExecution, createXai, mcpServer, viewImage, viewXVideo, webSearch, xSearch, xai, xaiTools };
+export { XaiRealtimeModel as Experimental_XaiRealtimeModel, type XaiRealtimeModelConfig as Experimental_XaiRealtimeModelConfig, VERSION, type XaiErrorData, type XaiFilesOptions, type XaiImageModelOptions, type XaiImageModelOptions as XaiImageProviderOptions, type XaiLanguageModelChatOptions, type XaiLanguageModelResponsesOptions, type XaiProvider, type XaiLanguageModelChatOptions as XaiProviderOptions, type XaiProviderSettings, type XaiLanguageModelResponsesOptions as XaiResponsesProviderOptions, type XaiSpeechModelOptions, type XaiTranscriptionModelOptions, type XaiVideoModelId, type XaiVideoModelOptions, type XaiVideoModelOptions as XaiVideoProviderOptions, codeExecution, createXai, mcpServer, viewImage, viewXVideo, webSearch, xSearch, xai, xaiTools };