English | မြန်မာဘာသာ | ဘာသာမန်
Mon is spoken by roughly one million people across Myanmar and Thailand. UNESCO classifies it as vulnerable — and no OCR toolchain existed for it before this project.
MonOCR takes an image of Mon script and returns text. It runs on Web, Android, and iOS — fully offline, no data leaves the device.
Built and maintained by the Mon developer community.
- Web: ocr.mondevhub.com
- Android: Google Play
- iOS: App Store (pending review)
Three real documents, the text the CLI returned for each, and the per-line records
behind it — nothing hand-corrected. See samples/, which also states
what was screened out and why the headline number is not an average.
Two of the three carry non-Unicode text layers, one Zawgyi and one legacy 8-bit, and both are among the cleanest results: rasterisation happens before the model, so no encoding ever reaches it.
All three apps ship one model, and since 2026-08-15 it is v3.5:
| Architecture | MobileNetV3-Large + SE + 2×BiLSTM-512 + attention + CTC |
| Parameters | 11,553,437 |
| Input | Grayscale, 160px height, static 1024px width |
| Charset | 276 characters, 277 classes |
| Precision | FP32 |
| Published at | janakhpon/monocr, revision d3d9d5e |
Android and iOS bundle it (46.2 MB and 46.3 MB respectively). The web app fetches it from that pinned revision. Per-app details are in apps/android, apps/ios and apps/web.
v3.5 is not a newer v2, it is a different contract. Input height went 128 to
160, output classes 316 to 277, charset 315 to 276, and the graph's width axis
went from dynamic to a static 1024. Anything still holding a cached v2 artifact
is refused rather than decoded, because a mismatch of that kind returns
well-formed Mon text that is wrong. v2 remains served at revision a51be11
for anyone pinned to it.
The model has no held-out evaluation. The figure that selected it is a training-time metric over 4,096 lines in a single typeface, and it is not an accuracy claim — see the model card.
A v4 server model was archived on 2026-08-05 under mon_OCR ADR-0011. It was
never trained to convergence, so archiving it was a decision about maintaining a
second path rather than about measured quality. It is not maintained.
No device latency number exists for any platform. Figures of that kind appeared here until 2026-08-15 and were architectural estimates, never measured on hardware.
Because high-quality Mon datasets are scarce, validated samples from the app's feedback flow feed directly into future training rounds.
The model deploys to Web, Android, and iOS — each using the format that enables hardware acceleration:
| Platform | Format | Acceleration |
|---|---|---|
| Web | ONNX | WASM |
| Android | ONNX | NNAPI |
| iOS | CoreML .mlpackage |
Apple Neural Engine |
- Web App — SvelteKit PWA
- Android App — Jetpack Compose
- iOS App — SwiftUI
- Feedback Service — Go ingestion API
- Shared Assets — model weights, locales, sync scripts
- HuggingFace — ONNX, CoreML, and checkpoint files
- npm package — JavaScript SDK
- Architecture decisions — ADRs
- API specs — OpenAPI contracts
- Mon Corpus Collection — training dataset
- Bugs: GitHub Issues
- Translations: Shared translation sheet
- Script samples: Contribute via the Android or iOS app, or reach out directly
- Standards: Contributing Guide · Security Policy