See where your AI budget goes.

Track every LLM call across OpenAI, Anthropic, and Google Gemini. One line of code. Full visibility.

import llmtracer
llmtracer.init()

# All OpenAI, Anthropic & Gemini calls are now tracked.

import llmtracer from '@llmtracer/sdk';
llmtracer.init();

// All OpenAI, Anthropic & Gemini calls are now tracked.

Start free → pip install llmtracer-sdk npm install @llmtracer/sdk

Cost Explorer Last 7 days $847.23 total

gpt-4o

$412.50

claude-sonnet-4-5

$203.18

gpt-4o-mini

$142.30

gemini-2.5-pro

$89.25

Source file	Cost	Calls
planning.py	$312.40	2,400
execution.py	$245.18	1,800
classifier.py	$89.20	4,200

All providers, one SDK

OpenAI, Anthropic, and Google Gemini. Auto-patches your SDK clients. Cost breakdown by model, source file, conversation, and user — automatically.

Full cost capture

Thinking tokens, cache tokens, tool tokens, reasoning tokens. Every billing dimension your provider charges for — captured automatically. Match your bill exactly.

Zero config

One line of code. Auto-flushes on exit. Captures caller file and function out of the box. No wrappers, no callbacks, no code changes.

The old way

× Install framework-specific adapter
× Create callback handler class
× Wire callbacks to every LLM call site
× Tag each call manually
× 8 files modified
× Hope you didn't miss one

With LLM Tracer