About Open Source AI Models
Open Source AI Modelsis an independent, curated directory of open-source large language models. Every entry on the site has weights you can actually download and run — no closed-API-only models, no "coming soon" placeholders.
We track parameters, context length, licence, benchmarks (MMLU, HumanEval, MATH and others), quantization options, VRAM requirements, hardware recommendations, and inference pricing across the major hosted providers. Every benchmark and pricing row carries a "last verified" date so you can see how fresh the data is.
The goal is simple: help engineers pick the right open-weights model on facts rather than marketing. The space moves fast and labs change their stories week to week — the directory aims to be the steady reference layer underneath all that.
What "open source" means here
We use the term broadly. Truly OSI-compliant open-source models (Apache 2.0, MIT) are listed alongside source-available models (Llama Community License, Gemma Terms of Use, Qwen License). The distinction matters for commercial use, so we flag it on every model and on each licence's detail page. See the licence comparison page for the breakdown.
Who builds this
Open Source AI Models is built and maintained by one person. The directory is updated as new releases ship — typically within a week of any major model announcement.
What we don't do
- We don't run benchmarks ourselves. All scores are from official model cards or papers.
- We don't take payment from labs or hosting providers to rank their models higher.
- We don't run advertising tracking. The site is read-only and doesn't collect personal data by default.
- We don't cover closed-API-only models. If you can't download the weights, it's not in the directory.
Open source
The site itself is open source. Source code lives at github.com/bryanflowers/opensourceaimodels. Pull requests welcome — particularly for adding missing models, fixing stale pricing, or refining licence summaries.