Analyze classical meter
Check Arabic verses against 16 classical meters with PyArud. Explore the صدر and عجز, binary scansion, feet, and the requested rawi.
LANGUAGE · POETRY · AI
Qafiya قَافِيَة.
An Arabic poetry workshop for classical meter, rhyme, readable scansion, and local vocalization.
Analysis is available in the live demo. AI generation is currently disabled.
A LOOK INSIDE
From a written verse to its rhythm.
BEHIND THE EXPERIENCE
From a written verse to its rhythm.
Complete vocalization locally
A bundled 4.8 MB libtashkeel ONNX model fills missing vowel marks on the CPU. Supplied vowels remain authoritative; analysis does not call an LLM.
Drafting is experimental
An optional provider adapter drafts poems and targets revisions at failed verses while preserving accepted text. Generation is currently disabled on the public demo.
BUILT WITH
The technology behind it.
Frontend: React 19, TypeScript, and Vite, with a responsive Arabic RTL interface.
Backend: Python 3.12, FastAPI, PyArud, NumPy, and CPU ONNX Runtime. A single worker shares a bounded inference session.
Hosting: Cloudflare Pages frontend and Render Docker API. GitHub Actions checks the frontend, backend, and production container.