
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 turns dense pages, tables, formulas and handwriting into clean structured text with a single compact 0.8B vision-language model.
All Tools
Discover our comprehensive AI-powered animation toolkit
MiniMax H3
MiniMax H3 AI Video Generator
Seedance 2.5
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
FLUX 3 Video Generator

Kling 3.0
Next-Gen AI Video Generator
Grok Video Generator
Create Videos from Text or Images with AI
MiniMax H3 video generator
MiniMax H3 AI Video Generator
What Makes Xiaomi OCR 0 Different
A single 0.8B checkpoint handles document parsing, OCR-centric question answering and key information extraction, replacing a whole stack of narrow single-purpose tools.
- One Compact 0.8B CheckpointDocument parsing, OCR-focused VQA and field extraction all live inside a single lightweight model.
- Benchmark-Leading AccuracyRanks first on Real5-OmniDocBench at 95.24 and takes second place on OmniDocBench v1.6 and its Wild variant.
- Reliable on Real-World ScansKeeps its accuracy on scanned, warped, skewed and photographed pages, not just clean digital PDFs.
Run Xiaomi OCR 0 in Three Simple Steps
From a single page image to clean, structured output in three quick steps.
Core Strengths of Xiaomi OCR 0
Compact footprint, robust parsing, handwriting coverage and open weights — the practical advantages this 0.8B OCR model delivers every day.
Full Document Parsing
Covers standard pages as well as scanned, warped, skewed and re-photographed documents.
Tables and Formulas
Reaches 97.84 on UniMER-Test for formulas, while table structure lands at 92.27 on PubTabNet.
Handwriting Recognition
Scores 73 on Chronicles-OCR for mature scripts, matching Seed 2.0 Pro.
OCR-Focused VQA
Averages 83.2 across five visual question answering suites, edging past a 2B rival at a fraction of the size.
Structured Field Extraction
Returns organized fields from receipts and forms rather than plain transcription.
Open Weights, Local Ready
Apache-2.0 weights plus GGUF and BF16 builds for llama.cpp, Ollama and Transformers.
Xiaomi OCR 0: Frequently Asked Questions
Straight answers about this compact OCR model, its benchmarks, its training recipe and local deployment.
What exactly is Xiaomi OCR 0?
It is a 0.8B vision-language checkpoint built specifically for OCR work — it reads documents and handles OCR-centric questions in one pass.
Which benchmarks does it lead?
It ranks first on Real5-OmniDocBench and second on both OmniDocBench v1.6 and Wild-OmniDocBench.
Does it handle handwriting?
Yes. Clerical, cursive, running and regular calligraphic scripts are all supported, scoring 73 on Chronicles-OCR.
How was the model trained?
Its recipe pairs roughly 170M OCR-focused samples with Q-Mask text anchoring, continued pretraining and mixed-task reinforcement learning.
Can I deploy it locally?
Yes. Apache-2.0 weights are public, with community GGUF and BF16 conversions for llama.cpp, Ollama and Transformers.
Does it only transcribe text?
No. It also answers questions about a page and pulls key fields into structured output.
Put Xiaomi OCR 0 to Work on Your Documents
Upload a page image and Xiaomi OCR 0 returns parsed text, tables and structured fields within seconds.
