Carrot Labs

SuperPenguin solves a real pain point: AI spend that spirals across multiple providers with no clear attribution to products or teams. The standout feature is per-request tracking via a two-line SDK — you tag calls with customer, feature, team, environment, and prompt version. This lets you see exactly which feature (e.g., 'support_agent') costs how much per customer. The Engineering ROI feature for Cursor and GitHub is uniquely valuable for dev teams using Cursor — it ties every dollar of Cursor spend to merged PRs, showing cost per engineer and project. The free tier (up to $2K managed spend) is generous and includes all dashboard features, SDK attribution, reconciliation, and email support. Growth ($30/mo, up to $5K spend) adds cost alerts and 3 team members. Pro ($200/mo, up to $20K spend) adds AI spend forecasting and Cursor PR cost tracking. Enterprise custom for $20K+ spend. We'd reach for this when you have multiple providers (OpenAI, Anthropic, Gemini, etc.) and need to attribute costs per customer/feature, or if you're a Cursor-heavy team wanting cost-per-PR. It's not a good fit if you just want a simple total spend dashboard — the free tier of any provider is enough. Non-technical teams may struggle with the SDK setup, though the dashboard works with just admin keys. Compared to alternative LLM cost tracking tools reviewed in June 2026, SuperPenguin positions as a pure cost management platform (not a gateway), focusing on granular attribution and reconciliation.