Documentation

Get started in two minutes

Install the wrapper, make a call, and see your first cost data. Your prompt content never reaches Tokeven — the wrapper sends only metadata.

Quickstart

1Install the wrapper

pip install tokeven

2Wrap your client

from anthropic import Anthropic
from tokeven import TokevenAnthropic

client = Anthropic()
tracked = TokevenAnthropic(client, ingest_token="tkv_ing_...")

3Use it exactly as before

response = tracked.messages.create(
    model="claude-sonnet-4-6",
    max_tokens=1024,
    messages=[{"role": "user", "content": "Hello"}]
)
# Cost, tokens, and model tracked automatically

Read the full setup guide

Browse the docs

Getting Started

Install, configure, and make your first tracked call

SDK Reference

Python wrapper API, configuration options, and advanced usage

Dashboards

Seven real-time dashboards for cost, models, tokens, patterns, drivers, and skills

Integrations

Connect Tokeven to your existing tools and workflows

API Reference

REST API for ingest, analytics, and export

Guides

Best practices for cost optimization and team rollout