SevenTnewS

OCR

4 published articles

AIFeatured3 min read

Analysis

The 0.8B model that just broke document parsing's pipeline monopoly

OvisOCR2 (0.8B parameters) from Tencent scored 96.58 on OmniDocBench, the first end-to-end model to top the leaderboard. The model uses reinforcement learning and distillation to outperform larger pipeline systems, challenging the necessity of multi-stage architectures.

2026-07-26

LLMs & ModelsFeatured3 min read

Document AI

Mistral's OCR 4 scores big, but its own audit shows why benchmark numbers don't tell the real story

Mistral OCR 4 introduces bounding boxes, block classification, and confidence scores alongside text extraction, supporting 170 languages. It achieves 72% human preference win rates and top benchmark scores, but Mistral's own analysis shows standard benchmarks penalize correct output for formatting artifacts, not accuracy errors.

2026-07-16

AI4 min read

Domain specialization

A six-month-old OCR model still beats Mistral. The reason is hard to fix.

DharmaOCR scores 0.925 on a Portuguese benchmark versus 0.798 for Mistral OCR4 and 0.7587 for Unlimited-OCR. The gap comes from concentrated training allocation and a DPO-based approach that suppresses text degeneration in complex documents.

2026-07-16

DeepSeek2 min read

Artificial Intelligence

DeepSeek-OCR 2 Brings Visual Causal Flow to Open-Source Document Understanding

DeepSeek-OCR 2 introduces Visual Causal Flow for human-like visual encoding. Now open-source on GitHub, the model supports vLLM and Transformers, dynamic resolution (up to 1,216 visual tokens), and document-to-markdown conversion.

2026-07-09