Developer Documentation

Build with Bear Lumen.

Track AI costs in one line of code.

Everything you need to instrument your AI app. Capture, attribute, and meter LLM spend across every provider, then query it from the SDK or the dashboard.

Quickstart

Install the SDK, wrap one LLM response, see real cost data in minutes.

SDKs

Node and Python libraries that auto-detect your provider.

REST API

Log events and query costs over raw HTTP, with examples.

Webhooks

Verify signatures and react to cost, billing, and sync events.

Quickstart

Your first tracked request

The SDK reads the response your provider already returned. No proxy, no added latency, no prompt payloads stored.

1
Install the SDK and add your API key.
2
Wrap the response your provider returned.
3
Query cost, margin, and usage anywhere.

track.ts

import { BearLumen } from '@bearlumen/node-sdk';

const bear = new BearLumen({
  apiKey: process.env.BEAR_LUMEN_API_KEY,
});

// Wrap the response you already have.
const result = bear.track(response);

result.costUsd;       // 0.0042
result.model;         // 'gpt-4o'
result.inputTokens;   // 12