ZeroLLM is an open-source directory, index, and health status dashboard for free Large Language Model (LLM) API providers. The project helps developers easily find, compare, and connect to free LLMs available across the web.
- Models Directory: Browse over 100+ free LLMs across multiple providers. Filter by modalities, status, and whether a credit card is required.
- Providers Directory: View and compare free tier limits, capabilities, and friction points across LLM API providers (e.g., Groq, OpenRouter, Google, Cohere, HuggingFace).
- Compare Models: Select up to 6 models side-by-side to compare context windows, output limits, rate limits, and capabilities.
- Config Generator: Generate drop-in configuration snippets for popular coding assistants and CLI tools like Cursor, Claude Code, and Codex.
- Playground: Test models instantly from the browser through a secure, non-logging local proxy server.
- Automated Health Checks: Daily cron jobs ping models to ensure they are online and responding correctly, automatically updating their statuses.
- Framework: Next.js 15 (App Router) + TypeScript
- Styling: Tailwind CSS + shadcn/ui
- Data Layer: Flat JSON structure updated automatically via sync scripts.
- Pipeline: Daily sync and health checks powered by GitHub Actions.
The easiest way to install and run ZeroLLM is by using the automated installation script. It will clone the repository, install dependencies, set up the environment, and create a global zerollm command.
Run the following command in your terminal:
curl -fsSL https://raw.githubusercontent.com/pisigmac/ZeroLLM/main/install.sh | bashOnce installed, you can start the dashboard from anywhere by running:
zerollm startOpen http://localhost:3000 in your browser to see the dashboard.
To update to the latest version in the future, simply run:
zerollm updateIf you prefer to set up the project manually:
git clone https://github.com/pisigmac/ZeroLLM.git
cd ZeroLLM
pnpm installCopy the example environment file and fill in any API keys you wish to use for the automated synchronization scripts.
cp .env.example .env.localpnpm devOpen http://localhost:3000 in your browser to see the dashboard.
The project includes two CLI scripts to maintain the accuracy of the models list and their health statuses.
- Sync Provider Data:
pnpm syncFetches the latest models from provider APIs and normalizes them into the local JSON data store. - Verify Model Health:
pnpm verifyPings all tracked models with a minimal 5-token request to verify they are online and functioning.
These scripts run automatically via GitHub Actions in production.
This project is licensed under the MIT License.