SuperPenguin documentation

Track and attribute AI spend with the SuperPenguin SDKs, or implement the Provider Billing API.

SuperPenguin gives developers one place to track AI cost per request, attribute spend to customers and features, and reconcile SDK estimates against provider invoices.

SuperPenguin observes SDK calls directly. Your traffic goes straight to the provider. The SDK is not a proxy.

Start here

Quickstart

Install an SDK, create an API key, wrap a client, and see attributed spend in minutes.

Attribution concepts

Learn how structured metadata turns raw usage into customer, feature, and team spend.

Python SDK

Wrap OpenAI, Azure OpenAI, Anthropic, Gemini, Bedrock, Deepgram, ElevenLabs, and LiteLLM.

TypeScript SDK

Native wrappers, Vercel AI SDK helpers, OpenTelemetry, and Realtime voice.

Setting up a company?

Create a company workspace

Choose an owner, select a plan, invite engineers, and configure shared access.

Set up AI Coding and PR Costs

See what Pro or Enterprise, Desktop, optional provider billing, and GitHub unlock.

Install SuperPenguin Desktop

Email the Mac app to engineers and begin syncing AI coding activity.

Plans and entitlements

Compare current prices, workspace limits, and feature availability.

Building a provider integration?

If you are an inference, speech, gateway, or agent provider that wants SuperPenguin to pull usage and billed costs from your API, start with the FOCUS AI Provider Profile.

Build a billing integration

Implement capabilities, usage, and cost exports, then run the conformance suite.

Customer teams connecting an existing provider account should use Connect an existing provider instead.