Look up context window size and per-token pricing for any AI model without visiting multiple provider docs
Build an app that lets users compare AI models by price, capability, and context length
Fetch all model metadata as a single JSON file to power a model-selection feature in an AI tool
Contribute a new AI model's specs by adding a TOML file and opening a pull request
| anomalyco/models.dev | arikchakma/maily.to | daltonmenezes/aura-theme | |
|---|---|---|---|
| Stars | 3,774 | 3,772 | 3,777 |
| Language | TypeScript | TypeScript | TypeScript |
| Setup difficulty | easy | moderate | easy |
| Complexity | 1/5 | 3/5 | 1/5 |
| Audience | developer | developer | developer |
Figures from each repo's GitHub metadata at analysis time.
Models.dev is a community-maintained open-source database that tracks AI models from different providers: what they can do, how much they cost, and what their technical limits are. The goal is to have one place where developers can look up information about any AI model rather than hunting through each provider's individual documentation. The data covers things like whether a model supports file attachments, tool calling, or chain-of-thought reasoning, what its pricing is per million tokens for input and output, how large a context window it has, when it was released, and whether its weights are publicly available. Each model's information is stored as a simple TOML file organized by provider inside the repository. The project exposes all this data through a public API: you can fetch a single JSON file containing information about every tracked model. Provider logos are also available as SVG files through the same API. The data is also used internally by opencode, an AI coding tool. Contributing is the main way the database stays current. Anyone can add a new provider or model by creating the appropriate folder and TOML file, then submitting a pull request. A GitHub Action automatically checks that submissions match the required schema before they can be merged. There is also a mechanism for reusing an existing model's definition when one provider is simply wrapping another provider's model, so data does not have to be duplicated. The project is written in TypeScript and is open source. It is designed to be useful to developers building applications that need to compare or select AI models programmatically.
A community-maintained open-source database of AI models from every major provider, listing their capabilities, pricing, context window sizes, and limits in one place so developers stop hunting through separate docs.
Mainly TypeScript. The stack also includes TypeScript, TOML.
Open source, check the repository for the specific license terms.
Setup difficulty is rated easy, with roughly 5min to a first successful run.
Mainly developer.
This repo across BitVibe Labs
Verify against the repo before relying on details.