A lightweight, open-source daemon that tracks AI API quotas across multiple providers in real time with a local dashboard.
onWatch is an open-source background daemon that monitors AI API quota usage across multiple providers like Synthetic, Z.ai, Anthropic, Codex, GitHub Copilot, MiniMax, Gemini CLI, and Antigravity. It solves the problem of limited visibility into API usage by providing real-time tracking, historical trends, and projections to help developers avoid throttling and manage budgets.
Developers and teams using AI coding assistants (e.g., Claude Code, Cline, GitHub Copilot) who need to monitor API usage across multiple providers. It's especially useful for solo developers, small teams, DevOps engineers, and privacy-conscious users in regulated industries.
Developers choose onWatch because it offers a unified, local monitoring solution with zero telemetry, a lightweight footprint (<50MB RAM), and support for eight major AI API providers—all in a single binary that's easy to deploy and self-update.
Track AI API quotas across Synthetic, Z.ai, Anthropic (Claude Code), Codex, GitHub Copilot & Antigravity in real time. Lightweight background daemon (<50MB RAM), SQLite storage, Material Design 3 dashboard. Zero telemetry.
Open-Awesome is built by the community, for the community. Submit a project, suggest an awesome list, or help improve the catalog on GitHub.
Aggregates quota data from eight AI API providers into a single Material Design 3 interface, offering a comprehensive view that individual provider dashboards lack.
Uses SQLite for data storage with zero telemetry, ensuring all historical data remains on the user's machine, which is ideal for regulated industries.
Consumes less than 50MB of RAM with all providers polling in parallel, making it efficient for continuous background operation.
Can update itself from the dashboard or CLI with automatic service restarts, reducing maintenance overhead for users.
Many features like notifications and menubar support are labeled as beta, leading to potential bugs, breaking changes, and limited reliability.
Setting up providers like GitHub Copilot requires specific tokens and scopes, and Docker deployments often need manual permission fixes due to non-root execution.
Only supports eight specific AI APIs; if a project relies on unsupported providers, onWatch cannot monitor them without custom development.