@ai-sdk/gateway 4.0.87 → 4.0.88

package/docs/00-ai-gateway.mdx

@@ -220,7 +220,7 @@ the failure without inspecting the Gateway's serialized error payload.
 </Note>
 
 AI Gateway supports durable text batches through `experimental_startBatch`,
-`experimental_getBatchStatus`, and `experimental_getBatchResults`. See
+`experimental_getBatchStatus`, `experimental_getBatchResults`, and `experimental_cancelBatch`. See
 [Batch](/docs/ai-sdk-core/batch) for the provider-agnostic workflow,
 and the [AI Gateway batch processing guide](https://vercel.com/docs/ai-gateway/models-and-providers/batch-processing)
 for supported models, limits, and Gateway-specific behavior.
@@ -414,6 +414,19 @@ Each request specifies its `type` and `model`. AI Gateway currently requires
 every text request in a batch to use the same model and throws before submission
 when the models differ.
 
+### Cancelling a batch
+
+Pass the batch reference returned by `experimental_startBatch` to request cancellation:
+
+```ts
+import { gateway } from '@ai-sdk/gateway';
+import { experimental_cancelBatch as cancelBatch } from 'ai';
+
+await cancelBatch({ provider: gateway, batch });
+```
+
+Cancellation is asynchronous. Continue checking `experimental_getBatchStatus` until the batch is terminal, then read `experimental_getBatchResults` for any partial results. A batch may finish before cancellation takes effect, and successful requests remain billable. Cancellation returns acknowledgement metadata, not confirmation that provider execution has stopped.
+
 ## Reranking Models
 
 You can create reranking models using the `rerankingModel` method on the provider instance:
@@ -736,7 +749,7 @@ import { generateText, tool } from 'ai';
 import { z } from 'zod';
 
 const { text } = await generateText({
-  model: 'xai/grok-4.6',
+  model: 'xai/grok-4.7',
   prompt: 'What is the weather like in San Francisco?',
   tools: {
     getWeather: tool({
@@ -1377,15 +1390,17 @@ The following gateway provider options are available:
 
   The unique identifier for the entity against which quota is tracked. Used for quota management and enforcement purposes.
 
-- **has** _Array&lt;'implicit-caching' | 'vision'&gt;_
+- **has** _Array&lt;'implicit-caching' | 'reasoning' | 'tool-use' | 'vision'&gt;_
 
   Restricts routing to provider models that have all of the specified capabilities. Applies to both BYOK and system credentials, since the capability is a property of the model rather than the credential. If no provider model for the requested model satisfies the capabilities, the request fails. Unsupported values are rejected.
 
   Supported capabilities:
   - `'implicit-caching'` — models that perform automatic (implicit) prompt caching.
+  - `'reasoning'` — models that support reasoning.
+  - `'tool-use'` — models that support tool calling.
   - `'vision'` — models that accept image input.
 
-  Example: `has: ['implicit-caching']` will only route to models that support implicit caching. Example: `has: ['vision']` will only route to models that accept image input.
+  Example: `has: ['implicit-caching']` will only route to models that support implicit caching. Example: `has: ['vision']` will only route to models that accept image input. Example: `has: ['tool-use']` will only route to models that support tool calling.
 
 - **providerTimeouts** _object_
 
@@ -1523,7 +1538,7 @@ const { text } = await generateText({
 
 #### Filtering by Model Capability
 
-Set `has` to restrict routing to provider models that have the specified capabilities. This applies to both BYOK and system credentials, since the capability is a property of the model rather than the credential. `'implicit-caching'` limits routing to models that perform automatic prompt caching, and `'vision'` limits routing to models that accept image input. If no provider model for the requested model satisfies the capabilities, the request fails.
+Set `has` to restrict routing to provider models that have the specified capabilities. This applies to both BYOK and system credentials, since the capability is a property of the model rather than the credential. `'implicit-caching'` limits routing to models that perform automatic prompt caching, `'vision'` limits routing to models that accept image input, `'reasoning'` limits routing to models that support reasoning, and `'tool-use'` limits routing to models that support tool calling. If no provider model for the requested model satisfies the capabilities, the request fails.
 
 ```ts
 import type { GatewayProviderOptions } from '@ai-sdk/gateway';

package/package.json

@@ -1,7 +1,7 @@
 {
   "name": "@ai-sdk/gateway",
   "private": false,
-  "version": "4.0.87",
+  "version": "4.0.88",
   "type": "module",
   "license": "Apache-2.0",
   "sideEffects": false,

package/src/gateway-batch.ts

@@ -2,6 +2,7 @@ import {
   InvalidArgumentError,
   UnsupportedFunctionalityError,
   type Experimental_BatchV4 as BatchV4,
+  type Experimental_BatchV4CancelResult as BatchV4CancelResult,
   type Experimental_BatchV4ItemResult as BatchV4ItemResult,
   type Experimental_BatchV4OperationOptions as BatchV4OperationOptions,
   type Experimental_BatchV4StartResult as BatchV4StartResult,
@@ -213,7 +214,54 @@ export class GatewayBatch implements BatchV4<{ text: GatewayModelId }> {
     }
   }
 
-  private getBatchUrl(path: 'results' | 'start' | 'status') {
+  /** Requests cancellation; status and partial results remain separate reads. */
+  async doCancelBatch({
+    batchId,
+    headers,
+    abortSignal,
+  }: BatchV4OperationOptions): Promise<BatchV4CancelResult> {
+    const resolvedHeaders = this.config.headers
+      ? await resolve(this.config.headers)
+      : undefined;
+
+    try {
+      const { value: responseBody } = await postJsonToApi({
+        url: this.getBatchUrl('cancel'),
+        headers: combineHeaders(
+          resolvedHeaders,
+          headers,
+          await resolve(this.config.o11yHeaders),
+        ),
+        body: { batchId },
+        successfulResponseHandler: createJsonResponseHandler(
+          gatewayBatchStatusResponseSchema,
+        ),
+        failedResponseHandler: createJsonErrorResponseHandler({
+          errorSchema: z.any(),
+          errorToMessage: data => getErrorMessage(data) ?? 'unknown error',
+        }),
+        ...(abortSignal && { abortSignal }),
+        fetch: this.config.fetch,
+      });
+
+      return {
+        ...(responseBody.providerMetadata != null && {
+          providerMetadata:
+            responseBody.providerMetadata as SharedV4ProviderMetadata,
+        }),
+      };
+    } catch (error) {
+      if (isAbortOrTimeoutError(error)) {
+        throw error;
+      }
+      throw await asGatewayError(
+        error,
+        await parseAuthMethod(resolvedHeaders ?? {}),
+      );
+    }
+  }
+
+  private getBatchUrl(path: 'cancel' | 'results' | 'start' | 'status') {
     return `${this.config.baseURL}/batch/${path}`;
   }
 }

package/src/gateway-language-model-settings.ts

@@ -29,6 +29,7 @@ export type GatewayModelId =
   | 'alibaba/qwen3.8-flash'
   | 'alibaba/qwen3.8-max'
   | 'alibaba/qwen3.8-max-0902'
+  | 'alibaba/qwen3.8-omni-flash'
   | 'amazon/nova-2-lite'
   | 'amazon/nova-lite'
   | 'amazon/nova-micro'
@@ -124,6 +125,7 @@ export type GatewayModelId =
   | 'mistral/mistral-medium-3.5'
   | 'mistral/mistral-nemo'
   | 'mistral/mistral-small'
+  | 'mixedbread/toast-1'
   | 'moonshotai/kimi-k2'
   | 'moonshotai/kimi-k2-thinking'
   | 'moonshotai/kimi-k2.5'
@@ -203,6 +205,8 @@ export type GatewayModelId =
   | 'perplexity/sonar-reasoning-pro'
   | 'poolside/laguna-s-2.1'
   | 'poolside/laguna-s-2.1-free'
+  | 'quiverai/arrow-2'
+  | 'quiverai/arrow-2-telos'
   | 'sakana/fugu-max'
   | 'sakana/fugu-ultra'
   | 'sakana/fugu-ultra-v2'
@@ -218,9 +222,11 @@ export type GatewayModelId =
   | 'spacexai/grok-4.3'
   | 'spacexai/grok-4.5'
   | 'spacexai/grok-4.6'
+  | 'spacexai/grok-4.7'
   | 'spacexai/grok-build-0.1'
   | 'stepfun/step-3.5-flash'
   | 'stepfun/step-3.7-flash'
+  | 'stepfun/step-5-preview'
   | 'tencent/hy-mt2-lite'
   | 'tencent/hy-mt2-plus'
   | 'tencent/hy-mt2-pro'
@@ -245,5 +251,6 @@ export type GatewayModelId =
   | 'zai/glm-5.3'
   | 'zai/glm-5.3-fast'
   | 'zai/glm-5.3-flash'
+  | 'zai/glm-5.3-flashx'
   | 'zai/glm-5v-turbo'
   | (string & {});

package/src/gateway-provider-options.ts

@@ -17,9 +17,10 @@ export type GatewayProviderOptions = {
 
   /**
    * Restrict routing to models that have all of the given capabilities.
-   * Currently supports `'implicit-caching'` and `'vision'` (image input).
+   * Currently supports `'implicit-caching'`, `'reasoning'`, `'tool-use'`, and
+   * `'vision'` (image input).
    */
-  has?: Array<'implicit-caching' | 'vision'>;
+  has?: Array<'implicit-caching' | 'reasoning' | 'tool-use' | 'vision'>;
 
   /**
    * Idempotency key for `experimental_startBatch`: retries with the same

package/src/gateway-speech-model-settings.ts

@@ -1,10 +1,7 @@
 export type GatewaySpeechModelId =
   | 'fish-audio/s1'
-  | 'fish-audio/s1-free'
   | 'fish-audio/s2-pro'
-  | 'fish-audio/s2-pro-free'
   | 'fish-audio/s2.1-pro'
-  | 'fish-audio/s2.1-pro-free'
   | 'openai/tts-1'
   | 'openai/tts-1-hd'
   | 'spacexai/grok-tts'

package/src/gateway-transcription-model-settings.ts

@@ -1,6 +1,5 @@
 export type GatewayTranscriptionModelId =
   | 'fish-audio/transcribe-1'
-  | 'fish-audio/transcribe-1-free'
   | 'google/gemini-3.5-transcribe'
   | 'google/gemini-3.5-transcribe-live'
   | 'openai/gpt-4o-mini-transcribe'