CHAMPOLLION
On rails, with guardrails.

Nasa beta na ngayon

Machine translation development and deployment on rails, built with low-resource and Indigenous language communities in mind.

We are putting the tools on rails: a project on your own computer that takes you from the files your community keeps to a measured model, while the people who hold the language decide where its sentences may go, what counts as good enough, and what ships. That release is being prepared; the rails promise you will know how good a model is — not that it is good. Today's releases are below.

Today's releases, on npm and PyPI — the CLI and the MCP server are source-available and free for noncommercial use; the evaluation harness is open source (AGPL):

Kinokopya ang setup prompt para sa inyong AI coding tool

O i-install ninyo mismo ang mga tool:

npm install champollionTranslation CLI
npx -y champollion-mcp-serverMCP server — para sa mga AI agent
pipx install mt-eval-harnessEvaluation harness
Mag-ambag ng compute sa project queue ?Ang queue po ay isang naka-rank na listahan ng mga translation benchmark na wala pang nagpapatakbo. Ini-install ng script ang harness, kinakalkula ang presyo ng trabaho at nagtatanong bago gumastos, pagkatapos ay pinapatakbo ang pinakataas ng queue gamit ang inyong sariling API key — hanggang sa budget na inyong itinakda — at pina-publish ang bawat resulta sa public board.
curl -fsSL https://champollion.dev/run_queue | bash -s -- --budget 2

Ang mismong site po ay nasa closed beta habang tinatapos namin ang muling pag-verify sa aming mga source, lisensya, at claim.

Magbubukas po ito kapag tapos na ang pagsusuri — at ginawa nang tama.

Para sa mga tanong, pagwawasto, o request para sa early access: info@champollion.dev. Lubos po naming tinatanggap ang mga pagwawasto sa anumang cited figure.

Hindi po nakilala ang key na iyon.