None
None
Long Story Video Skill
Turn your idea into 10s–10min video—AI writes scripts, prompts, and generates footage automatically.
Ads Video Skill
Generate professional ads and sales videos—AI auto-generates scripts, prompts, and footage.
3D Science Explainer Video Skill
Convert scientific concepts into stunning 3D explain animations
Feedback
freeTrialImage.bannerPity
freeTrialImage.upgradeUnlock
- ✓freeTrialImage.benefitHd
- ✓freeTrialImage.benefitWatermark
- ✓freeTrialImage.benefitUnlimited
xiaomi ocr 0
Xiaomi OCR 0 converts scans, tables, formulas and handwriting into clean structured text — free, no signup, with results in seconds.
All Tools
Discover our comprehensive AI-powered animation toolkit
Muse Video
Seedance2.5
The Future of AI Video Is Here.
Seedance 2.0
The Future of AI Video Is Here.
Veo3.1
Create Stunning Videos with Veo3.1
Grok Video Generator
Create Videos from Text or Images with AI
MiniMax H3 video generator
MiniMax H3 AI Video Generator
Seedance 2.0
The Future of AI Video Is Here.
Gemini Omni 1.2 Flash
One Compact OCR Engine, Many Document Jobs
The Xiaomi OCR 0 checkpoint handles page parsing, OCR-focused Q&A and key-field extraction in one pass, replacing several narrow utilities with a single compact model.
- All-in-One 0.8B CheckpointPage parsing, OCR-driven visual Q&A and field extraction share a single compact checkpoint.
- Benchmark-Leading AccuracyIt ranks first on Real5-OmniDocBench with 95.24 and places second on OmniDocBench v1.6 and Wild-OmniDocBench.
- Reliable on Real-World ScansPerforms well on scanned, warped, skewed and camera-captured pages, not just pristine PDFs.
From Page Image to Clean Output in Three Steps
Upload, prompt and export — a short workflow turns any page image into structured text.
What This 0.8B OCR Model Delivers
Compact size, tough document handling, handwriting recognition and open weights define the practical strengths of this OCR model.
Robust Document Parsing
Covers ordinary pages as well as scanned, warped, skewed and re-photographed documents.
Tables and Formulas
Formula recognition hits 97.84 on UniMER-Test, and table structure reaches 92.27 on PubTabNet.
Handwriting and Calligraphy
Reaches 73 on Chronicles-OCR mature scripts, matching Seed 2.0 Pro.
OCR-Focused Visual Q&A
Averages 83.2 across five visual Q&A suites, surpassing a 2B competitor at a fraction of the size.
Structured Field Extraction
Extracts structured fields from receipts and forms rather than stopping at plain transcription.
Open Weights, Local Ready
Apache-2.0 weights and GGUF/BF16 builds support llama.cpp, Ollama and Transformers.
Questions About This 0.8B OCR Model
Answers on accuracy, handwriting, training and running this compact model locally.
What exactly is this model?
It is a 0.8B vision-language checkpoint made for OCR tasks, reading pages and answering OCR-focused questions in a single pass.
Where does it rank on benchmarks?
It holds first place on Real5-OmniDocBench and second on OmniDocBench v1.6 and Wild-OmniDocBench.
Does it support handwriting?
Yes — clerical, cursive, running and regular scripts are handled, scoring 73 on Chronicles-OCR.
What training approach was used?
The recipe combines about 170M OCR-centric samples, Q-Mask text anchoring, continued pretraining and mixed-task reinforcement learning.
Can it run on local hardware?
Yes — Apache-2.0 weights and community GGUF/BF16 builds work with llama.cpp, Ollama and Transformers.
Is transcription its only job?
No — it also answers questions about a page and pulls key fields into structured output.
Put This OCR Model to Work on Your Pages
Send in a page image and receive parsed text, tables and structured fields within seconds — no setup required.