Vercel AI SDK First class

Scrawn wraps the Vercel AI SDK so every streamText, generateText, streamObject, and generateObject call auto-bills on every step finish. No manual token counting, no per-route instrumentation.

Setup

Pass the Vercel AI SDK module to biller.ai() with your pricing config:

import * as ai from "ai";
import { scrawn, mul, inputTokens, outputTokens } from "@scrawn/core";

const biller = scrawn({
  apiKey: process.env.SCRAWN_KEY,
  baseURL: process.env.SCRAWN_BASE_URL,
});

const aii = biller.ai(ai, {
  inputDebit: mul(biller.tag("GPT_INPUT"), inputTokens()),
  outputDebit: mul(biller.tag("GPT_OUTPUT"), outputTokens()),
});

The returned aii object has the same API as the Vercel AI SDK — call streamText, generateText, etc. exactly as you normally would, just with a userId parameter added.

Usage

Every wrapped function works exactly like the original Vercel AI SDK — just add a userId to associate billing:

const result = await aii.streamText({
  userId: "user-123",             // ← added for billing
  model: openai("gpt-4o-mini"),
  prompt: "Hello",
});

The same pattern applies to generateText, streamObject, and generateObject — all original options and types are preserved.

Per-Model Pricing

Use tags to charge different rates per model:

const biller = scrawn({
  apiKey: process.env.SCRAWN_KEY,
  baseURL: process.env.SCRAWN_BASE_URL,
  tags: [
    "GPT_INPUT", "GPT_OUTPUT",
    "CLAUDE_INPUT", "CLAUDE_OUTPUT",
    "LLAMA_INPUT", "LLAMA_OUTPUT",
  ] as const,
});

const aii = biller.ai(ai, {
  inputDebit: biller.tag("GPT_INPUT"),
  outputDebit: biller.tag("GPT_OUTPUT"),
});

Manual Tracking (trackAI)

For full manual control without the biller.ai() wrapper, use trackAI() inside onFinish or onStepFinish:

import { mul, inputTokens, outputTokens } from "@scrawn/core";

ai.streamText({
  model: openai("gpt-4o"),
  prompt: "Hello",
  onFinish: async (event) => {
    await biller.trackAI(
      "user-123",
      event.model,
      {
        inputTokens: event.usage.inputTokens ?? 0,
        outputTokens: event.usage.outputTokens ?? 0,
        totalTokens: event.usage.totalTokens ?? 0,
      },
      {
        inputDebit: mul(biller.tag("GPT_INPUT"), inputTokens()),
        outputDebit: mul(biller.tag("GPT_OUTPUT"), outputTokens()),
      },
      {
        inputDebit: 3,
        outputDebit: 6,
        inputCacheDebit: 3,
        outputCacheDebit: 6,
      }
    );
  },
});

Compatibility

The wrapper supports ai SDK versions 3.x and 4.x. If you encounter type issues, ensure ai and @scrawn/core are both on recent versions.