# MiniCPM-OCR Jev checks OCR outputs for reliability, flagging low-confidence results

> Original title: Thanks for the great breakdown! 💙🙌 Reliable document AI isn't just about extracting information, but also knowing when to double-check ...

- Source: OpenBMB
- Published: 2026-11-05T01:40:49.000Z
- Bad Influence: https://badinfluence.news/items/hllcsm7rq2wn7evwdbj4xxooq
- Original: https://x.com/OpenBMB/status/2118156551072710709

## Summary

OpenBMB releases MiniCPM-OCR Jev, a model that scores the reliability of OCR outputs, including extracted text, formulas, and tables. The companion MiniCPM-OCR parser uses 1.3B parameters to extract information from Chinese and English documents. Jev identifies potentially unreliable results for human review but does not correct them.

## Post

Thanks for the great breakdown! 💙🙌 Reliable document AI isn't just about extracting information, but also knowing when to double-check it. That's exactly what we're aiming for with MiniCPM-OCR and Jev! 🚀

## Quoted post @Olaemmy1200

> OCR pipelines can be much more reliable when you add a verification model.
> 
> Instead of trusting every extracted result, you can use a second model to check how accurate the output actually is.
> 
> A good example is MiniCPM OCR from @OpenBMB and ModelBest.
> 
> It combines a lightweight document parser with a separate model, MiniCPM OCR Jev, designed to evaluate OCR reliability.
> 
> Here’s how it works:
> 
> 1. Document parsing
> 
> MiniCPM-OCR uses just 1.3B parameters to extract information from Chinese and English documents.
> 
> It handles scanned documents, phone photos, curved pages, tilted inputs, and even key information extraction.
> 
> 2. Output verification
> 
> MiniCPM-OCR Jev evaluates the extracted text, formulas, and tables to estimate how reliable the results are.
> 
> You can also use it to verify outputs from other OCR models, meaning you don’t have to replace your existing pipeline.
> 
> Why does this matter for AI agents?
> 
> Agents are increasingly working with PDFs, invoices, research papers, and other structured documents.
> 
> But one incorrect number or poorly extracted table can affect every step that follows.
> 
> A simple workflow would look like this:
> 
> → Parse the document with MiniCPM-OCR.
> 
> → Pass the extracted content to Jev for reliability scoring.
> 
> → Route low-confidence results to a human reviewer or another correction step.
> 
> One important distinction: Jev identifies potentially unreliable outputs. It doesn’t automatically correct them.
> 
> The project also provides open weights, fine tuning support, and inference compatibility with llama.cpp, vLLM, SGLang, and Ollama.
> 
> There are also integrations through Agent Skills.
> 
> If you’re building document processing agents, this is worth testing.
> 
> Especially on difficult inputs like dense tables, mathematical formulas, scanned pages, and low quality phone photos.
> 
> The bigger idea: Don’t just build agents that extract information. Build pipelines that can also evaluate whether that information should be trusted.

https://x.com/Olaemmy1200/status/2117919541065810107