Contribute

Add your project.

OpenOdia lists the whole Odia open-source ecosystem, not one maintainer's work. If you have built something for ଓଡ଼ିଆ — a model, a dataset, a font, a keyboard, a library, an app — it belongs here.

The template

These are the fields the directory actually renders. An entry that arrives complete is citable and filterable from day one.

- [Project name](https://github.com/owner/repo) : One sentence on what it does and who it is for (Apache-2.0).

# Fields the directory reads:
#   name         required   what it is called
#   url          required   GitHub repo, Hugging Face model/dataset, or project page
#   description  required   one sentence, plain English, no marketing
#   license      required   SPDX id in parentheses, e.g. (MIT), (Apache-2.0), (CC-BY-4.0)
#   task         required   what it does: ASR, TTS, translation, OCR, NER, corpus, font, keyboard…
#   size         datasets   rows or hours, e.g. 120k sentence pairs / 40 hours audio
#   citation     optional   BibTeX, if the project has a paper

What gets listed

It is about Odia

The project targets ଓଡ଼ିଆ specifically, or covers it as a named language alongside others. A general multilingual tool with no Odia support is out of scope.

It is open

Source, weights, or data are publicly available under a stated license. A license is required — an entry without one is listed as “No license”, which is a warning to researchers, not a badge.

It resolves

The link points at something live: a GitHub repo, a Hugging Face model or dataset, or a project page that still exists.

It says what it is

One plain sentence. Descriptions that read as marketing copy get rewritten before listing.