SuperPenguin

Carrot Labs AI, Inc. builds SuperPenguin, an AI spend intelligence platform that attributes what a company spends on AI to the customer, feature, team or pull request that caused it. Providers report a total bill; SuperPenguin breaks that bill down so teams can see which parts of the product and which engineering work are driving cost, catch spikes before the invoice, and forecast the month, without sitting in the path of any request. It is for engineering and finance teams running AI in production and paying for AI coding tools. It works in three ways. Attribution uses Python and TypeScript SDKs to tag API spend with your own dimensions, with no cardinality caps. Coding ROI connects to GitHub and uses a free Mac app on each engineer's machine to put a cost on every merged pull request across Cursor, Claude Code and Codex. One View pulls every provider into a single total with breakdowns by provider, model, project and API key, and reconciles against invoices. Supported sources include OpenAI, Anthropic, Gemini, Deepgram, ElevenLabs, Vercel, AWS Bedrock, Azure, Modal, LiteLLM, Together AI, Fireworks AI, OpenRouter and Cursor. Pricing is based on the AI spend under management, not seats or requests. Free is $0 for up to $2,000 of managed spend, one member and attribution for three merged pull requests in total. Growth is $30 per month for up to $5,000, three members, Slack, email and Discord alerts, and a 90 day backfill. Pro is $200 per month for up to $20,000, ten members, unlimited pull request attribution and a month end forecast. Enterprise is custom for spend above $20,000. The listing name is the company; the product is marketed under the SuperPenguin brand.