ModelBest and OpenBMB Open-Source MiniCPM-OCR Series, Delivering High-Performance Document Parsing and AI-Powered Quality Control
Original HTML
ModelBest and OpenBMB Open-Source MiniCPM-OCR Series, Delivering High-Performance Document Parsing and AI-Powered Quality Control
BEIJING, Oct. 31, 2026 /PRNewswire/ -- ModelBest and OpenBMB jointly announced the release and open-source availability of the MiniCPM-OCR series, a family of models for efficient document parsing and parsing-quality assessment. The launch includes two models: MiniCPM-OCR 1.0 (1.3B) and MiniCPM-OCR Jev (2B).
As the first OCR-focused model in the MiniCPM family, MiniCPM-OCR 1.0 brings top-tier parsing accuracy to a compact 1.3-billion-parameter architecture optimized for edge deployment, pushing forward the balance between efficiency and performance for open-source OCR models. Built for real-world conditions, it accurately parses scanned, photographed, curved, tilted and screen-captured documents without cumbersome image pre-correction, lowering the barrier to large-scale document digitization.
The model handles dense bilingual Chinese-English content, complex mathematical formulas, and multi-level or borderless tables, consistently performing at a leading level across a broad range of mainstream industry evaluations. Beyond full-page reconstruction, it supports key information extraction (KIE), pulling specified fields from invoices, IDs and financial statements according to business instructions.
Powered by DSpark decoding acceleration, MiniCPM-OCR 1.0 significantly increases end-to-end throughput while lowering per-page latency. For enterprises processing millions of pagesâsuch as financial contract archives, academic libraries and government recordsâthis means processing far more documents with the same hardware and time budget, at a substantially lower computing cost.
To address the long-standing need for page-by-page manual review, ModelBest and OpenBMB also introduced MiniCPM-OCR Jev, the first OCR result decision model that acts as an "AI quality inspector." Rather than performing OCR itself, Jev compares the source image with OCR output and rapidly assigns confidence scores to text, LaTeX formulas and tables, detecting missing characters, errors and formatting misalignments. Despite its compact 2B-parameter size, Jev delivers judgment accuracy typically associated with far larger models.
Jev supports cross-model deployment and can be integrated into existing OCR workflows without replacing the underlying OCR engine. It can automatically identify reliable results for direct processing, flag potentially inaccurate outputs for human review and select the best result from multiple OCR systems. This enables enterprises to move from full manual inspection toward risk-based, automated quality control.
The MiniCPM-OCR series is available on GitHub, Hugging Face and ModelScope, and supports vLLM, SGLang, llama.cpp and Ollama, along with SDKs, APIs, Agent Skill plugins and open-source fine-tuning code. Day-0 hardware adaptation covers AMD, Arm, Houmo AI, Intel and NVIDIA. Through the FlagOS software stack, it also has Day-0 support for AI and edge chips from CIX Technology, Huawei Ascend, Hygon, Iluvatar CoreX, Kunlunxin, MetaX, Moore Threads, Sunrise, T-Head, TsingMicro, NVIDIA and Apple Mac (Arm).
SOURCE ModelBest
Share this article
Original Summary
<p>BEIJING, Oct. 31, 2026 /PRNewswire/ -- ModelBest and OpenBMB jointly announced the release and open-source availability of the MiniCPM-OCR series, a family of models for efficient document parsing and parsing-quality assessment. The launch includes two models: MiniCPM-OCR 1.0 (1.3B) and...</p>
Generated Article_json
{"target_audience": "b2b", "is_medical": false, "ai_generated_title": "ModelBest, OpenBMB Open-Source 1.3B MiniCPM-OCR Model and 2B Quality Checker", "body_html": "<p><strong>BEIJING – [CURRENT_DATE] --</strong> ModelBest and OpenBMB have open-sourced the MiniCPM-OCR series, two models for document parsing and parsing-quality assessment: MiniCPM-OCR 1.0 (1.3 billion parameters) and MiniCPM-OCR Jev (2 billion parameters).</p><h5>MiniCPM-OCR 1.0 parses distorted documents without image pre-correction</h5><p>MiniCPM-OCR 1.0 is the first OCR-focused model in the MiniCPM family. Its compact architecture is optimized for edge deployment. It parses scanned, photographed, curved, tilted and screen-captured documents without image pre-correction, which lowers the barrier to large-scale digitization.</p><p>The model handles dense bilingual Chinese-English content, complex mathematical formulas, and multi-level or borderless tables. It also supports key information extraction (KIE), pulling specified fields from invoices, IDs and financial statements according to business instructions.</p><h5>DSpark decoding acceleration raises throughput for million-page workloads</h5><p>DSpark decoding acceleration increases end-to-end throughput and lowers per-page latency. For enterprises processing millions of pages, such as financial contract archives, academic libraries and government records, the same hardware and time budget covers more documents at lower computing cost.</p><h5>Jev scores OCR output to replace page-by-page manual review</h5><p>MiniCPM-OCR Jev is described as the first OCR result decision model. It does not perform OCR. It compares the source image with OCR output and assigns confidence scores to text, LaTeX formulas and tables. It detects missing characters, errors and formatting misalignments. ModelBest and OpenBMB say the 2B-parameter model delivers judgment accuracy typically associated with far larger models.</p><p>Jev can be integrated into existing OCR workflows without replacing the underlying engine. It identifies reliable results for direct processing, flags doubtful outputs for human review and selects the best result from multiple OCR systems. Enterprises can therefore move from full manual inspection to risk-based, automated quality control.</p><h5>Release covers GitHub, Hugging Face and ModelScope with day-0 support for 14 chip platforms</h5><p>The series is available on GitHub, Hugging Face and ModelScope. It supports vLLM, SGLang, llama.cpp and Ollama, and ships with SDKs, APIs, Agent Skill plugins and open-source fine-tuning code.</p><p>Day-0 hardware adaptation covers AMD, Arm, Houmo AI, Intel and NVIDIA. Through the FlagOS software stack, day-0 support extends to chips from CIX Technology, Huawei Ascend, Hygon, Iluvatar CoreX, Kunlunxin, MetaX, Moore Threads, Sunrise, T-Head and TsingMicro, plus NVIDIA and Apple Mac (Arm).</p>", "meta_description": "ModelBest and OpenBMB open-source MiniCPM-OCR 1.0 (1.3B) and Jev (2B), adding edge-ready document parsing and automated OCR quality control with broad chip support.", "seo_keywords": "MiniCPM-OCR, ModelBest, OpenBMB, open-source OCR, document parsing, OCR quality control, key information extraction, edge AI, FlagOS", "companies": ["ModelBest", "OpenBMB"], "industries": ["Information & Communication Technology (ICT)"], "markets_regions": ["China", "Global"], "news_type": ["product_launch"], "products": ["MiniCPM-OCR"], "themes": ["open_source_ai", "document_digitization", "quality_control"], "events": [], "cta_text": "Access the MiniCPM-OCR series on GitHub, Hugging Face or ModelScope", "learn_more": "", "fact_check": {"accuracy_score": 100, "legal_risk": "low", "corrections_made": [], "recommendation": "publish"}}
Content Hash (SHA-256, erste 32 hex)
37d448b684168a5c5f15c8f8e747f0a1