Vercel AI SDK

Trace Vercel AI SDK calls with axonPushMiddleware. The middleware hooks into the SDK’s experimental middleware API to capture LLM start, end, and streaming token events.

Installation

npm install @axonpush/sdk ai

Setup

import { AxonPush } from "@axonpush/sdk";
import { axonPushMiddleware } from "@axonpush/sdk/integrations/vercel-ai";

const client = new AxonPush({
  apiKey: process.env.AXONPUSH_API_KEY!,
  tenantId: process.env.AXONPUSH_TENANT_ID!,
});

const middleware = axonPushMiddleware({
  client,
  channelId: 1,
  agentId: "my-agent",
});

Usage

import { generateText } from "ai";
import { openai } from "@ai-sdk/openai";

const { text } = await generateText({
  model: openai("gpt-4o"),
  prompt: "Explain quantum computing.",
  experimental_middleware: middleware,
});

Events Traced

Event	When
llm.start	A model call begins
llm.end	A model call completes
llm.token	A streaming token is received