Evidence-first authority page
Open-Source AI Projects Built in India: How to Evaluate One
What to check on an open-source AI project's own repository before adopting it, rather than a ranked list of Indian open-source AI.
Direct answer
This site does not maintain a ranked list of "best" open-source AI projects from India, because openness itself is a spectrum — license terms, release cadence, and who actually maintains the code vary project to project and change over time. The reliable check is the project's own repository, not a static summary of it.
Last reviewed 2026-09-22
Reviewed by Febin Francis, Founder/Editor on 2026-09-22 under the methodology-only publication path: this page makes no externally-verifiable factual or comparative claim, so it is reviewed for originality and disclosed limitations rather than sourced against a primary receipt.
What "open-source" doesn't tell you by itself
A project can publish weights without publishing training data, publish code under a restrictive license, or stop receiving commits entirely while still being described as open-source somewhere online. None of that is visible from a one-line description — it's visible in the repository's license file, commit history, and issue tracker.
For Indian-language AI specifically, also check whether a project's claimed language coverage has been evaluated on held-out data for that language, or only mentioned in a launch post. The two are not the same, and this site's own glossary entries on instruction tuning and benchmark contamination explain why that distinction matters.
Where to look instead of a summary page
- The repository's license file, for what you can actually do with the code or weights.
- Commit and release history, for whether the project is still maintained.
- The models directory on this site for Indian model identities with a source-linked primary reference.