SuperPenguin
SuperPenguin tracks AI provider spend and attributes it to customers, features, teams, and pull requests.
The Problem
Teams using multiple AI providers (OpenAI, Anthropic, Gemini, coding tools like Cursor and Claude Code) get a separate bill from each vendor, and those bills only show totals by provider, project, or API key. In the example shown, only $15,658 of a $22,478 monthly bill could be mapped to a customer, leaving the rest unaccounted for. Companies cannot see what a given customer, feature, team, or merged pull request actually costs in AI spend, and provider dashboards cannot group spend across vendors or tie it to business dimensions. This makes it hard to catch cost spikes before the invoice arrives or to judge whether a feature or customer is profitable.
The Solution
SuperPenguin is an AI spend intelligence platform that pulls every provider bill into one total and lets teams break it down by customer, feature, team, or pull request. Users add a Python or TypeScript SDK and pass metadata (like a customer id) on each API call so spend is tracked per request without sitting in the request path, adding under 10ms. A free Mac app syncs per-session cost data from Cursor, Claude Code, and Codex on a developer's machine, and connecting GitHub lets the product price each merged pull request (Coding ROI). It supports OpenAI, Anthropic, Gemini, AWS Bedrock, Azure, Deepgram, ElevenLabs, Together AI, Fireworks AI, Modal, OpenRouter, Vercel AI Gateway, LiteLLM, and Cursor, and offers spend alerts via Slack, Discord, or email along with end-of-month forecasting on higher tiers.