@ai-sdk/xai 4.0.35 → 4.0.37

package/CHANGELOG.md

@@ -1,5 +1,18 @@
 # @ai-sdk/xai
 
+## 4.0.37
+
+### Patch Changes
+
+- a4d386d: feat(xai): add the Grok 4.6 model IDs and support its `xhigh` reasoning effort
+
+## 4.0.36
+
+### Patch Changes
+
+- 8edc775: feat (provider/xai): support Grok Imagine Video 1.5. Adds the `grok-imagine-video-1.5` model id and native `1080p` for text-to-video and image-to-video (the standard `resolution: '1920x1080'` now maps to `1080p`). Reference-to-video remains capped at `720p`, so a `1080p` request in that mode is downgraded with a warning. Also fixes reference routing: previously any non-empty `inputReferences` array selected reference-to-video, so an array holding only a non-image reference sent `reference_images: []` with no usable references.
+- 3d05053: feat (provider/xai): add `referenceVoiceIds` for reference-to-video reference audio. Pass up to 3 xAI preset voice ids (e.g. `['eve']`) and reference them from the prompt with `<AUDIO_0>`–`<AUDIO_2>`; they are sent as `reference_audios: [{ voice_id }]` on `POST /v1/videos/generations`.
+
 ## 4.0.35
 
 ### Patch Changes

package/dist/index.d.ts

@@ -3,13 +3,14 @@ import * as _ai_sdk_provider_utils from '@ai-sdk/provider-utils';
 import { InferSchema, FetchFunction, WebSocketConstructor } from '@ai-sdk/provider-utils';
 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-4.5' | 'grok-latest' | (string & {});
+type XaiChatModelId = 'grok-4.20-non-reasoning' | 'grok-4.20-reasoning' | 'grok-4.3' | 'grok-4.5' | 'grok-4.6' | 'grok-latest' | (string & {});
 declare const xaiLanguageModelChatOptions: z.ZodObject<{
     reasoningEffort: z.ZodOptional<z.ZodEnum<{
         none: "none";
         low: "low";
         medium: "medium";
         high: "high";
+        xhigh: "xhigh";
     }>>;
     logprobs: z.ZodOptional<z.ZodBoolean>;
     topLogprobs: z.ZodOptional<z.ZodNumber>;
@@ -72,7 +73,7 @@ declare const xaiFilePartProviderOptions: z.ZodObject<{
 }, z.core.$strip>;
 type XaiFilePartProviderOptions = z.infer<typeof xaiFilePartProviderOptions>;
 
-type XaiResponsesModelId = 'grok-4.20-non-reasoning' | 'grok-4.20-reasoning' | 'grok-4.3' | 'grok-4.5' | 'grok-latest' | (string & {});
+type XaiResponsesModelId = 'grok-4.20-non-reasoning' | 'grok-4.20-reasoning' | 'grok-4.3' | 'grok-4.5' | 'grok-4.6' | 'grok-latest' | (string & {});
 /**
  * @see https://docs.x.ai/docs/api-reference#create-new-response
  */
@@ -82,6 +83,7 @@ declare const xaiLanguageModelResponsesOptions: z.ZodObject<{
         low: "low";
         medium: "medium";
         high: "high";
+        xhigh: "xhigh";
     }>>;
     reasoningSummary: z.ZodOptional<z.ZodEnum<{
         auto: "auto";
@@ -115,11 +117,12 @@ declare const xaiImageModelOptions: z.ZodObject<{
 }, z.core.$strip>;
 type XaiImageModelOptions = z.infer<typeof xaiImageModelOptions>;
 
-type XaiVideoModelId = 'grok-imagine-video' | (string & {});
+type XaiVideoModelId = 'grok-imagine-video' | 'grok-imagine-video-1.5' | (string & {});
 
 declare const resolutionSchema: z.ZodEnum<{
     "480p": "480p";
     "720p": "720p";
+    "1080p": "1080p";
 }>;
 type XaiVideoResolution = z.infer<typeof resolutionSchema>;
 interface XaiVideoSharedOptions {
@@ -156,6 +159,10 @@ interface XaiVideoReferenceToVideoOptions extends XaiVideoSharedOptions, XaiVide
     mode: 'reference-to-video';
     /** Reference image URLs (1-7) for R2V generation. */
     referenceImageUrls: string[];
+    /**
+     * Preset voice ids (up to 3) that give the subject a voice.
+     */
+    referenceVoiceIds?: string[];
 }
 interface XaiVideoGenerationOptions extends XaiVideoSharedOptions, XaiVideoUserOptions {
     mode?: undefined;
@@ -177,6 +184,10 @@ interface XaiLegacyReferenceToVideoOptions extends XaiVideoSharedOptions, XaiVid
      */
     mode?: undefined;
     referenceImageUrls: string[];
+    /**
+     * Preset voice ids (up to 3) that give the subject a voice.
+     */
+    referenceVoiceIds?: string[];
 }
 /**
  * Provider options for xAI video generation.

package/dist/index.js

@@ -305,13 +305,14 @@ var xaiLanguageModelChatOptions = z2.object({
    * - `low` (default): Uses some reasoning tokens, but still fast.
    * - `medium`: More thinking for less-latency-sensitive applications.
    * - `high`: Uses more reasoning tokens for deeper thinking.
+   * - `xhigh`: Uses the most reasoning tokens (supported by `grok-4.6`).
    *
    * Note: Not every Grok model accepts every value. Refer to xAI's docs for
    * the values supported by your selected model.
    *
    * @see https://docs.x.ai/docs/guides/reasoning
    */
-  reasoningEffort: z2.enum(["none", "low", "medium", "high"]).optional(),
+  reasoningEffort: z2.enum(["none", "low", "medium", "high", "xhigh"]).optional(),
   logprobs: z2.boolean().optional(),
   topLogprobs: z2.number().int().min(0).max(8).optional(),
   /**
@@ -557,7 +558,7 @@ var XaiChatLanguageModel = class _XaiChatLanguageModel {
             low: "low",
             medium: "medium",
             high: "high",
-            xhigh: "high"
+            xhigh: this.modelId === "grok-4.6" ? "xhigh" : "high"
           },
           warnings
         });
@@ -1966,11 +1967,12 @@ var xaiLanguageModelResponsesOptions = z8.object({
    * Constrains how hard a reasoning model thinks before responding.
    * Possible values are `none` (disables reasoning entirely; supported by
    * `grok-4.3` and newer reasoning models), `low` (uses fewer reasoning
-   * tokens), `medium`, and `high` (uses more reasoning tokens).
+   * tokens), `medium`, `high` (uses more reasoning tokens), and `xhigh`
+   * (supported by `grok-4.6`).
    *
    * @see https://docs.x.ai/docs/guides/reasoning
    */
-  reasoningEffort: z8.enum(["none", "low", "medium", "high"]).optional(),
+  reasoningEffort: z8.enum(["none", "low", "medium", "high", "xhigh"]).optional(),
   reasoningSummary: z8.enum(["auto", "concise", "detailed"]).optional(),
   logprobs: z8.boolean().optional(),
   topLogprobs: z8.number().int().min(0).max(8).optional(),
@@ -2461,7 +2463,7 @@ var XaiResponsesLanguageModel = class _XaiResponsesLanguageModel {
             low: "low",
             medium: "medium",
             high: "high",
-            xhigh: "high"
+            xhigh: this.modelId === "grok-4.6" ? "xhigh" : "high"
           },
           warnings
         });
@@ -3687,7 +3689,7 @@ var xaiTools = {
 };
 
 // src/version.ts
-var VERSION = true ? "4.0.35" : "0.0.0-test";
+var VERSION = true ? "4.0.37" : "0.0.0-test";
 
 // src/files/xai-files.ts
 import {
@@ -3812,53 +3814,18 @@ import { z as z20 } from "zod/v4";
 import { lazySchema as lazySchema8, zodSchema as zodSchema8 } from "@ai-sdk/provider-utils";
 import { z as z19 } from "zod/v4";
 var nonEmptyStringSchema = z19.string().min(1);
-var resolutionSchema = z19.enum(["480p", "720p"]);
+var resolutionSchema = z19.enum(["480p", "720p", "1080p"]);
 var modeSchema = z19.enum(["edit-video", "extend-video", "reference-to-video"]);
 var baseFields = {
   pollIntervalMs: z19.number().positive().nullish(),
   pollTimeoutMs: z19.number().positive().nullish(),
   resolution: resolutionSchema.nullish()
 };
-var userField = {
-  user: z19.string().optional()
-};
-var editVideoSchema = z19.object({
-  ...baseFields,
-  ...userField,
-  mode: z19.literal("edit-video"),
-  videoUrl: nonEmptyStringSchema,
-  referenceImageUrls: z19.undefined().optional()
-});
-var extendVideoSchema = z19.object({
-  ...baseFields,
-  mode: z19.literal("extend-video"),
-  videoUrl: nonEmptyStringSchema,
-  referenceImageUrls: z19.undefined().optional()
-});
-var referenceToVideoSchema = z19.object({
-  ...baseFields,
-  ...userField,
-  mode: z19.literal("reference-to-video"),
-  referenceImageUrls: z19.array(nonEmptyStringSchema).min(1).max(7),
-  videoUrl: z19.undefined().optional()
-});
-var autoDetectSchema = z19.object({
-  ...baseFields,
-  ...userField,
-  mode: z19.undefined().optional(),
-  videoUrl: nonEmptyStringSchema.optional(),
-  referenceImageUrls: z19.array(nonEmptyStringSchema).min(1).max(7).optional()
-});
-var xaiVideoModelOptions = z19.union([
-  editVideoSchema,
-  extendVideoSchema,
-  referenceToVideoSchema,
-  autoDetectSchema
-]);
 var runtimeSchema = z19.looseObject({
   mode: modeSchema.optional(),
   videoUrl: nonEmptyStringSchema.optional(),
   referenceImageUrls: z19.array(nonEmptyStringSchema).min(1).max(7).optional(),
+  referenceVoiceIds: z19.array(nonEmptyStringSchema).max(3).optional(),
   user: z19.string().optional(),
   ...baseFields
 });
@@ -3868,6 +3835,7 @@ var xaiVideoModelOptionsSchema = lazySchema8(
 
 // src/xai-video-model.ts
 var RESOLUTION_MAP = {
+  "1920x1080": "1080p",
   "1280x720": "720p",
   "854x480": "480p",
   "640x480": "480p"
@@ -3885,35 +3853,39 @@ function resolveStartImage(options) {
   return (_a = getFirstFrameImage(options)) != null ? _a : options.image;
 }
 var isVideoFile = (file) => file.mediaType != null && getTopLevelMediaType3(file.mediaType) === "video";
-function fileToXaiImageUrl(file) {
+var isImageReference = (file) => file.mediaType == null || getTopLevelMediaType3(file.mediaType) === "image";
+function fileToXaiUrl(file) {
   if (file.type === "url") {
     return file.url;
   }
   const base64Data = typeof file.data === "string" ? file.data : convertUint8ArrayToBase64(file.data);
   return `data:${file.mediaType};base64,${base64Data}`;
 }
-function resolveReferenceImages(options, xaiOptions, warnings) {
+function resolveReferences(options, xaiOptions, warnings) {
   if (options.inputReferences != null && options.inputReferences.length > 0) {
-    const imageReferences = options.inputReferences.filter((reference) => {
-      if (isVideoFile(reference)) {
+    const imageFiles = [];
+    for (const reference of options.inputReferences) {
+      if (!isImageReference(reference)) {
         warnings.push({
           type: "unsupported",
           feature: "inputReferences",
-          details: 'xAI reference-to-video accepts image references only. The video reference was ignored. Use providerOptions.xai.mode "extend-video" to continue from a video.'
+          details: isVideoFile(reference) ? 'xAI reference-to-video accepts image references only. The video reference was ignored. Use providerOptions.xai.mode "extend-video" to continue from a video.' : "xAI reference-to-video accepts image references only. The non-image reference was ignored."
         });
-        return false;
+        continue;
       }
-      return true;
-    });
-    return imageReferences.map((reference) => ({
-      url: fileToXaiImageUrl(reference)
-    }));
+      imageFiles.push(reference);
+    }
+    return imageFiles.length > 0 ? imageFiles.map((reference) => ({ url: fileToXaiUrl(reference) })) : void 0;
   }
   if ((xaiOptions == null ? void 0 : xaiOptions.referenceImageUrls) != null && xaiOptions.referenceImageUrls.length > 0) {
     return xaiOptions.referenceImageUrls.map((url) => ({ url }));
   }
   return void 0;
 }
+function hasImageInputReference(options) {
+  var _a, _b;
+  return (_b = (_a = options.inputReferences) == null ? void 0 : _a.some(isImageReference)) != null ? _b : false;
+}
 function resolveVideoMode(options, xaiOptions) {
   if ((xaiOptions == null ? void 0 : xaiOptions.mode) != null) {
     return xaiOptions.mode;
@@ -3922,9 +3894,8 @@ function resolveVideoMode(options, xaiOptions) {
     return "edit-video";
   }
   const hasFrameImages = options.frameImages != null && options.frameImages.length > 0;
-  const hasInputReferences = options.inputReferences != null && options.inputReferences.length > 0;
   const hasLegacyReferenceUrls = (xaiOptions == null ? void 0 : xaiOptions.referenceImageUrls) != null && xaiOptions.referenceImageUrls.length > 0;
-  if (!hasFrameImages && (hasInputReferences || hasLegacyReferenceUrls)) {
+  if (!hasFrameImages && (hasImageInputReference(options) || hasLegacyReferenceUrls)) {
     return "reference-to-video";
   }
   return void 0;
@@ -4029,7 +4000,7 @@ var XaiVideoModel = class {
         warnings.push({
           type: "unsupported",
           feature: "resolution",
-          details: `Unrecognized resolution "${options.resolution}". Use providerOptions.xai.resolution with "480p" or "720p" instead.`
+          details: `Unrecognized resolution "${options.resolution}". Use providerOptions.xai.resolution with "480p", "720p", or "1080p" instead.`
         });
       }
     }
@@ -4049,7 +4020,7 @@ var XaiVideoModel = class {
           details: 'xAI does not accept a video as a start/frame image. The video was ignored. Use providerOptions.xai.mode "extend-video" to continue from a video instead.'
         });
       } else {
-        body.image = { url: fileToXaiImageUrl(startImage) };
+        body.image = { url: fileToXaiUrl(startImage) };
       }
     }
     const lastFrameImage = getLastFrameImage(options);
@@ -4061,20 +4032,50 @@ var XaiVideoModel = class {
       });
     }
     if (hasReferenceImages) {
-      const referenceImages = resolveReferenceImages(
-        options,
-        xaiOptions,
-        warnings
-      );
+      const referenceImages = resolveReferences(options, xaiOptions, warnings);
       if (referenceImages != null) {
         body.reference_images = referenceImages;
+      } else {
+        warnings.push({
+          type: "unsupported",
+          feature: "referenceImages",
+          details: "xAI reference-to-video requires at least one image reference. The video will be generated without reference images."
+        });
+      }
+      const referenceVoiceIds = xaiOptions == null ? void 0 : xaiOptions.referenceVoiceIds;
+      if (referenceVoiceIds != null && referenceVoiceIds.length > 0) {
+        body.reference_audios = referenceVoiceIds.map((voiceId) => ({
+          voice_id: voiceId
+        }));
+      }
+      if (body.resolution === "1080p") {
+        warnings.push({
+          type: "unsupported",
+          feature: "resolution",
+          details: "xAI reference-to-video is limited to 720p. The request was downgraded from 1080p to 720p."
+        });
+        body.resolution = "720p";
       }
     }
+    if (body.resolution === "1080p" && this.modelId === "grok-imagine-video") {
+      warnings.push({
+        type: "unsupported",
+        feature: "resolution",
+        details: 'xAI model "grok-imagine-video" does not support 1080p. Use "grok-imagine-video-1.5" for 1080p, or a lower resolution. The request was sent with 1080p.'
+      });
+    }
     if (options.inputReferences != null && options.inputReferences.length > 0 && !hasReferenceImages) {
       warnings.push({
         type: "unsupported",
         feature: "inputReferences",
-        details: "xAI only supports inputReferences for reference-to-video generation. The reference images were ignored."
+        details: hasImageInputReference(options) ? "xAI only supports inputReferences for reference-to-video generation. The reference images were ignored." : "xAI reference-to-video requires at least one image reference. The references were ignored."
+      });
+    }
+    if ((xaiOptions == null ? void 0 : xaiOptions.referenceVoiceIds) != null && xaiOptions.referenceVoiceIds.length > 0 && !hasReferenceImages) {
+      warnings.push({
+        type: "unsupported",
+        feature: "referenceVoiceIds",
+        details: "xAI only supports reference voices for reference-to-video generation. The reference voices were ignored."
       });
     }
     if (!isExtension && (xaiOptions == null ? void 0 : xaiOptions.user) !== void 0) {
@@ -4089,6 +4090,7 @@ var XaiVideoModel = class {
           "resolution",
           "videoUrl",
           "referenceImageUrls",
+          "referenceVoiceIds",
           "user"
         ].includes(key)) {
           body[key] = value;