DigestAI news desk

Cut through the AI noise.

Listen
Developing story · 2 episodes · 2 Oct to 15 Oct

Datalab Launches New AI Document Benchmarks

Datalab is advancing its open-source tools for evaluating document processing accuracy by releasing new benchmarks. The latest development is the launch of OmniParseBench, an extensive OCR benchmark containing over sixteen thousand tests.

  1. Researchnew

    Datalab releases OmniParseBench, an open OCR benchmark with 16,288 tests

    Datalab has released OmniParseBench, an open-source benchmark designed to evaluate Optical Character Recognition (OCR) parsers. The dataset consists of 16,288 pass/fail tests derived from 2,937…

    1 source
  2. Researchnew

    Datalab releases OmniExtractBench to audit PDF extraction accuracy

    Datalab launched OmniExtractBench, an open benchmark for structured document extraction from PDFs into JSON. It pools 620 documents from four existing datasets, including regulatory filings and…

    1 source
Who and what
DatalabMarkTechPostLlamaIndexGPT-5.6-solLlamaExtractGeminiClaudeReducto deepextract v2Mistral OCR 4.1Gemini 3.8 FlashClaude Sonnet 5.5Reducto
hot Character.AI files for Chapter 11 bankruptcy, Disney offers $20 million for assets new Meta survey finds 76% of APAC small businesses report AI productivity gains hot OpenAI's GPT-6 Astra: Changing Work and Research new Google launches Gemini 4 Argon model globally for Pro and Ultra subscribers with introductory API pricing hot LFM2.5-24B-A2B: Built for fast on-device agents new Washington AG files notice in Character.AI bankruptcy over child safety hot OpenAI engineer charged with bugging executive conference room new Meta Muse AI agent designs RISC-V crypto SoC for $2,000