@mastra/ai-sdk

The recommended way of using Mastra and AI SDK together is by installing the @mastra/ai-sdk package. @mastra/ai-sdk provides custom API routes and utilities for streaming Mastra agents in AI SDK-compatible formats. Including chat, workflow, and network route handlers, along with utilities and exported types for UI integrations.

Installation

bash

npm install @mastra/ai-sdk

Usage

If you want to use dynamic agents you can use a path with :agentId.

typescript

import { chatRoute } from '@mastra/ai-sdk';

export const mastra = new Mastra({
  server: {
    apiRoutes: [
      chatRoute({
        path: '/chat/:agentId',
      }),
    ],
  },
});

Or you can create a fixed route (i.e. /chat):

typescript

import { chatRoute } from '@mastra/ai-sdk';

export const mastra = new Mastra({
  server: {
    apiRoutes: [
      chatRoute({
        path: '/chat',
        agent: 'weatherAgent',
      }),
    ],
  },
});

After defining a dynamic route with :agentId you can use the useChat() hook like so:

typescript

type MyMessage = {};

const { error, status, sendMessage, messages, regenerate, stop } = useChat<MyMessage>({
  transport: new DefaultChatTransport({
    api: 'http://localhost:4111/chat/weatherAgent',
  }),
});

chatRoute() forwards the incoming request AbortSignal to agent.stream(). If the client disconnects, Mastra aborts the in-flight generation. If you need generation to continue and persist server-side after disconnect, build a custom route around agent.stream(), avoid passing the request signal, and call consumeStream() on the returned MastraModelOutput.

Workflow route

Stream a workflow in AI SDK-compatible format.

typescript

import { workflowRoute } from '@mastra/ai-sdk';

export const mastra = new Mastra({
  server: {
    apiRoutes: [
      workflowRoute({
        path: '/workflow',
        agent: 'weatherAgent',
      }),
    ],
  },
});

Network route

Stream agent networks (routing + nested agent/workflow/tool executions) in AI SDK-compatible format.

typescript

import { networkRoute } from '@mastra/ai-sdk';

export const mastra = new Mastra({
  server: {
    apiRoutes: [
      networkRoute({
        path: '/network',
        agent: 'weatherAgent',
      }),
    ],
  },
});

Framework-agnostic handlers

For use outside the Mastra server (e.g., Next.js App Router, Express), you can use the standalone handler functions directly. These handlers return a compatibility ReadableStream that can be passed to AI SDK response helpers like createUIMessageStreamResponse and pipeUIMessageStreamToResponse:

handleChatStream

typescript

import { handleChatStream } from '@mastra/ai-sdk';
import { createUIMessageStreamResponse } from 'ai';
import { mastra } from '@/src/mastra';

export async function POST(req: Request) {
  const params = await req.json();
  const stream = await handleChatStream({
    mastra,
    agentId: 'weatherAgent',
    params,
  });
  return createUIMessageStreamResponse({ stream });
}

handleWorkflowStream

typescript

import { handleWorkflowStream } from '@mastra/ai-sdk';
import { createUIMessageStreamResponse } from 'ai';
import { mastra } from '@/src/mastra';

export async function POST(req: Request) {
  const params = await req.json();
  const stream = await handleWorkflowStream({
    mastra,
    workflowId: 'myWorkflow',
    params,
  });
  return createUIMessageStreamResponse({ stream });
}

handleNetworkStream

Pass AI SDK UIMessage[] from your installed ai version so TypeScript can infer the correct stream overload.

Handlers keep the existing v5/default behavior. If your app is typed against ai@6, pass version: 'v6'.

typescript

import { handleNetworkStream } from '@mastra/ai-sdk';
import { createUIMessageStreamResponse, type UIMessage } from 'ai';
import { mastra } from '@/src/mastra';

export async function POST(req: Request) {
  const params = (await req.json()) as { messages: UIMessage[] };
  const stream = await handleNetworkStream({
    mastra,
    agentId: 'routingAgent',
    version: 'v6',
    params,
  });
  return createUIMessageStreamResponse({ stream });
}

Manual transformation

If you have a raw Mastra stream, you can manually transform it to AI SDK UI message parts:

Use toAISdkStream for both versions. If your app is typed against ai@6, pass version: 'v6'.

typescript

import { toAISdkStream } from '@mastra/ai-sdk';
import { createUIMessageStream, createUIMessageStreamResponse } from 'ai';

export async function POST(req: Request) {
  const { messages } = await req.json();
  const agent = mastra.getAgent('weatherAgent');
  const stream = await agent.stream(messages);

  // deduplicate messages https://ai-sdk.dev/docs/troubleshooting/repeated-assistant-messages
  const uiMessageStream = createUIMessageStream({
    originalMessages: messages,
    execute: async ({ writer }) => {
      for await (const part of toAISdkStream(stream, { from: 'agent' })) {
        writer.write(part);
      }
    },
  });

  return createUIMessageStreamResponse({ stream: uiMessageStream });
}

For AI SDK v6, select the v6 stream contract explicitly:

typescript

const uiMessageStream = createUIMessageStream({
  originalMessages: messages,
  execute: async ({ writer }) => {
    for await (const part of toAISdkStream(stream, {
      from: 'agent',
      version: 'v6',
    })) {
      writer.write(part);
    }
  },
});