Say it with your hands.

SignFlow reads your camera, recognizes the ASL alphabet on-device, and helps you spell words letter by letter.

Best in Chrome or Safari with a front camera. Nothing ever leaves your device.

Now spelling

Fingerspelling, spelled back to you — the avatar page shows every hand shape.

What it recognizes

the full ASL alphabet — 24 static letters plus the motion-based J and Z, recognized by tracing the fingertip path (manual buttons remain as a fallback) — plus single-shape word signs like I LOVE YOU, YES, and NO, and anything you teach it on the Teach page.

The avatar can also sign back to you — fingerspelling with natural letter flow. Browse common words in the sign library, or drill them in word practice: watch a word being signed, then type what you saw.

This is a fingerspelling assistant, not a full ASL translator: it does not recognize grammar, facial expressions, or continuous signing.

Private by construction

Camera frames are processed entirely on your device via WebAssembly. No video or images are uploaded, recorded, or logged — ever. Transcripts stay in memory unless you save them, and you can wipe everything in Settings.

Honest limitations

Recognition depends on lighting, angle, and how closely your hand shapes match the training data (one contributor's hand). It works best for deliberate, paused fingerspelling — not continuous signing. Some letters (M/N/T, I/Y) occasionally confuse. J and Z are traced as movements — they need a deliberate finger path, so a sloppy or fast trace may miss (manual buttons remain). Built-in word signs (I LOVE YOU, YES, NO) are approximate single hand shapes — the real signs add motion, so treat them as hints and learn from the library's references.