Models.dev Builds the Open-Source Database Tracking Every AI Model's Pricing

models.dev is a community-built, open-source database of AI model specs, pricing, and capabilities — queryable via a free JSON API.

·
·
·
Models.dev Builds the Open-Source Database Tracking Every AI Model's PricingPRO
Read2 min
TypeRepo
SubtopicDatasets
  • Models.dev is a free, open-source database of AI model specs, pricing, and capabilities across dozens of providers.
  • Data is stored as TOML files in a public GitHub repo and served via a free JSON API at models.dev/api.json.
  • Each model entry tracks token costs, context limits, modalities, feature flags (tool calling, reasoning, structured output), and lifecycle status.
  • Built and used internally by the opencode team (SST/Anomaly Innovations), which supports 75+ providers via the registry.
  • Contributing requires only a TOML file and a PR; a GitHub Action validates schema automatically.
  • Coverage depends on community contributions — pricing and new models can lag; not a guaranteed billing source of truth.

Models.dev is a community-driven, open-source database that tracks AI model specifications, pricing, and capabilities across dozens of providers. It's free to use, free to contribute to, and already has 6,400+ GitHub stars and 1,500+ forks. If you've ever had to manually cross-reference context window sizes, token costs, or feature flags across OpenAI, Anthropic, Google, Mistral, and a dozen others, this project exists specifically to end that pain.

The problem it solves

There's no single database with information about all available AI models, so the team started Models.dev as a community-contributed project to address this. The AI model landscape has exploded: dozens of providers, hundreds of model variants, constantly shifting pricing, and wildly inconsistent documentation. Keeping that up to date in any single application is a full-time job. Models.dev externalizes that work to the community.

Models.dev aggregates AI model specifications, pricing information, and capabilities across multiple providers, serving model metadata through a public JSON API and an interactive web interface, with data stored in TOML files validated by Zod schemas and deployed to Cloudflare's edge network.

How it works

The architecture is deliberately simple. Model data lives in plain TOML files inside the GitHub repo, organized by provider and model. A build step validates every file against a strict schema, then generates the JSON endpoints that power the website and API. The system processes TOML configuration files through validation and aggregation stages before serving them via a Cloudflare Worker.

Pro article

This story is for Pro members

You've reached the end of the free preview. Upgrade to AlphaSignal Pro to read the full article - and everything else behind the paywall.

Trending
  • No trending articles

Comments

avatar

Next Reads