TL;DR: One encoder and one JSON schema turn documents, images, audio and video into structured, grounded text. It posts the top reported score on document parsing. Youtu-Parsing-Omni: One Encoder, One Schema, All Modalities Key points ・🧩 A single shared Transformer encodes images, audio and video (initialized from Qwen3-0.6B-Base) ・⏱ Rotary positions split into time/height/width, so video and sound share one timeline ・📐 OmniSchema puts 10 tasks (documents, chemistry, sheet music, charts, audio, video...) into one JSON format ・🔁 Post-training combines OSAD distillation from a self-refreshing teacher with schema-aware RLVR ・📄 OmniDocBench v1.6: 96.96 overall, the highest reported ・🎬 OmniParsingBench: 75.08, best among open-weight models; text-rich video 75.62 vs 65.36 Gemini-3-Pro still scores higher at 77.44, and it trails Qwen3-Omni on Video-MME-v2. ASR is also a weak spot. Still, a promising step toward an "any-input parser" front end for LLM agents. #OmniModal# #DocumentParsing#

A table that's quietly misaligned. A paragraph that's silently dropped. Ever had a document parse that looked right but wasn't? Title: OCR Is Dead. Long Live Agentic OCR URL: Drawing on lessons from building LlamaParse, LlamaIndex lays out what it means to treat document parsing as agentic: the system decides how to read a document, calls tools or specialized models as needed, and changes approach based on results. Three highlights stand out. 🧩 Adaptive reading strategy Instead of processing every element uniformly, resources are allocated based on document complexity — tables get specialized attention while simpler text gets standard handling. 🔍 Evidence-based self-correction Simply reprocessing a problematic section isn't enough. The system has to verify, with checkable evidence, that a correction actually improves accuracy without sacrificing fidelity. ⏱️ Clear limits and objectives Explicit constraints on time and cost coexist with a single goal: faithfully reconstructing the entire document, down to headers, footers, and other page furniture. What stands out here is the bar itself moving — once AI agents rely on parsed documents for consequential work, "readable" isn't enough; it has to be faithfully reconstructed. #AIAgents# #DocumentParsing#

We comprehensively evaluated 16 recent frontier VLMs - including Opus 5.5 and GPT-6 Sol/Luna - on whether higher effort led to better document parsing performance. Higher effort typically leads to improvements on other benchmarks (coding, knowledge work), but up until recently it wasn't obvious that this natively improved capabilities for reading PDFs. Results: ✅ Out of the frontier models, Opus 5.5 has the best performance relative to its price. It's especially good at parsing tables. ✅ Astra is also quite good, but starts at a more expensive price than Opus. ✅ GPT-6 Luna is more compelling at the cheaper end of doc parsing If you're parsing documents at scale, you'll still want a dedicated OCR solution like LlamaParse ( that has better performance at a cheaper price. But if you're parsing docs "in the agent loop" within an app like Codex/Claude code, and you're too lazy to integrate a dedicated solution, then Opus 5.5 is currently the leader. Full results on ParseBench:

Alibaba has open-sourced Logics-Parsing-V3, a 0.8B VLM for structured long-document parsing. Instead of resetting at every page, it carries a compact structural state forward as it reads — preserving context across pages without putting the entire document into one context window. This lets it recover document hierarchy, merge content split across pages, and link visual elements to related text in a single end-to-end model. It also handles complex layouts, scientific formulas, and chemical notation. On MPDocBench-Parse, it scores 85.26, ranking #1# — 4.46 points ahead of PaddleOCR-VL-1.5

Two new frontier models just landed on Databricks. @OpenAI GPT-6.1 Sol and @SpaceXAI Grok 4.7 are both live today. GPT-6.1 Sol leads the cost-quality Pareto frontier on OfficeQA Pro v2. Grok 4.7 reaches the frontier at enterprise document parsing. Two benchmark leaders, available the day they ship. Frontier models are shipping faster than ever. Unity Gateway keeps you right there with them. It picks the right model for the right task, governs every call, and controls costs across GPT-6.1 Sol, Grok 4.7, and 60+ other frontier and open models already on Databricks. Try GPT-6.1 Sol and Grok 4.7 today:

Cohere Parse 5 scores 79.2 on ParseBench — ahead of Mistral OCR 4, Databricks AI Parse, and Azure Document Intelligence. Best price-performance on the market for high-volume document parsing. The enterprise data pipeline just got a new default.

.@Google's latest model, Gemini 3.7 Flash, is now available on Databricks! Gemini 3.7 Flash delivers significant improvements in document parsing over version 3.6, and it now leads among Gemini models on our internal Databricks coding benchmark. Run Gemini 3.7 Flash where your data already lives - governed, secure, and ready for custom AI apps and agents built with the data in your Lakehouse. Govern every call, monitor performance, and scale securely across your AI apps with Unity AI Gateway:

Unlimited-OCR is now on ModelScope! A 3.3B multilingual OCR model for one-shot parsing across single images, multi-page documents, and PDFs. License: MIT🚀 🤖 📄 Full-document parsing instead of cropped-region OCR 📏 32K output length for long OCR sequences 🧩 Base and gundam image modes for different document layouts ⚙️ Transformers inference + SGLang serving with OpenAI-compatible streaming requests Built to push DeepSeek-OCR-style document parsing further.

Exciting day for Databricks customers! We’re launching three new frontier models on Databricks today. ↳ @OpenAI's GPT-6 Sol and GPT-6 Luna, with strong performance at lower cost and a higher Pareto frontier on the OfficeQA Pro benchmark ↳ @AnthropicAI's Claude Opus 5.5, with substantial improvements in cost efficiency and strong performance on enterprise document parsing tasks The industry’s best models are shipping faster than ever. Putting them into production and scaling your AI footprint means governing every model call, controlling costs, monitoring performance, and using the right model for the right task. Unity Gateway gives you one governance layer across every model, so you can govern and audit requests, set budgets, track costs, and choose the best model for each job.