Bayān: Touch Grass with open-source AI and local Arabic reflection

Description and use

Bayān is a local-first Arabic rhetoric companion that encourages users to step away from screens, observe the environment, and use local AI to craft stronger Arabic expressions.

Problem

Users want a low-friction way to practice Arabic imagery, sensory language, and thoughtful observation without relying on cloud AI services or complex interfaces.

Outcome

Bayān provides a minimal, privacy-preserving, offline-capable experience that turns a short outdoor walk into a focused reflective writing task, supporting Arabic imagery and sensory language.

Workflow

• Select a rhetorical exercise (simile, metaphor, or sensory description).
• Step outside and observe a detail in the environment.
• Return to the app and use the local AI to transform the observation into a stronger Arabic phrase. (Optional: use local speech synthesis or transcription.)
• Write or record the final reflection. (Fallback behavior ensures a coherent experience if the model is unavailable.)

Evidence

Bayān is a local-first Arabic rhetoric companion that gets people off the screen and back into the world.

Source

The project uses: Ollama with Gemma 3 4B for mission generation, translation, and chat; FastAPI for the backend service layer; a lightweight static web frontend for the game flow; browser-local storage for lightweight game state; optional local Piper speech synthesis for Arabic playback; optional local Whisper transcription for voice-note input.

Source

The app also includes validation and fallback behavior so that if the local model is unavailable or returns weak output, the user still receives a coherent experience instead of a broken UI.

Source