Tag
A new vision-capable model called Clef reportedly matches Jev's capabilities but adds image support, enabling document understanding tasks like checking signatures or validating invoice totals with calibrated results. The author questions why the release has been under-hyped despite astounding results.
LightOn 发布新一代轻量级端到端 OCR 模型 LightOnOCR-3 系列，其中 4B 版本采用 Qwen3.5 视觉-语言架构，新增 grounding（带边界框的布局标签）、图像描述和图表数据提取等视觉理解能力，全部以 Apache 2.0 许可开源。
The ICDAR2026 competition on multimodal reasoning over documents presented results from 8 teams, highlighting advanced systems that use structured evidence extraction and multi-component orchestration for VQA tasks across diverse document domains.
This paper introduces TestHallVQA, a multi-image VQA benchmark for evaluating Large Vision-Language Models' document-level reasoning under redundant contexts, and proposes a new metric F1-R2 to quantify computational reasoning and robustness.
Elon Musk announces that Grok 4.6 reached #1 on Databricks, with Ivan Zhou reporting SOTA performance on OfficeQA Pro V2 using Databricks's Genie harness.
A user shares surprising results testing Liquid AI's new LFM2.5-VL-3B vision-language model across many languages, noting strong visual capabilities but weaker instruction following; Liquid AI announces the model can read screens, documents, and ground objects to coordinates.
An empirical attribution study of failures in multi-page visually rich document understanding (MP-VRDU), isolating representation, selection, and reasoning failure modes and providing guidance for building such systems under fixed compute budgets.
NVIDIA releases Nemotron Parse 2.0, a document image parsing model that converts scanned PDFs and images into structured text with layout, bounding boxes, and reading order, adding multilingual OCR improvements and chart-aware parsing.
Introduces XL-DocBench, a human-verified benchmark for extra-long document understanding with 1,519 questions across six professional domains, requiring multi-page evidence and structured reasoning, showing current LLMs still struggle with long-context professional documents.
This paper presents VLD-RAG, an agentic multimodal retrieval-augmented generation framework for question answering over long, visually-rich documents. It uses a page-preserving index and a verifier-guided agent workflow to improve cross-page evidence retrieval and reasoning, outperforming prior vision-based baselines on benchmarks like LongDocURL and MMLongBench-Doc.
This paper reveals that Vision-Language Models often rewrite rather than faithfully transcribe text when encountering perturbations like typos or visual artifacts, introducing the FaithC4 benchmark to evaluate this behavior across multiple models and languages.
This tweet benchmarks Gemini 3.6 Flash and Gemini 3.5 Flash Lite on document understanding, finding that while the Flash series initially excelled at visual understanding, recent versions have plateaued or regressed due to posttraining for coding and reasoning.
DeepSeek OCR 2 is a 3B parameter model that outperforms Gemini 3 Pro in OCR and document understanding, featuring a human-like reading order and support for local fine-tuning.
SynthDocBench is a fully synthetic benchmark for long-context visual document understanding that systematically controls document length, layout, modality, and question type, revealing failure modes in current VLMs such as length degradation and positional sensitivity.
LlamaIndex benchmarked GPT-5.6 on document understanding and found no improvement over GPT-5.5; the model performs well on text and tables but struggles with charts and layout.
Introduces BaFCo, a benchmark dataset for Bangla form comprehension focusing on Document Layout Analysis (DLA) and Key Information Extraction (KIE). It includes 200 multi-page complex Bangladeshi government forms with fine-grained annotations across 26 entity types and evaluates multiple MLLMs, revealing limitations in understanding complex Bangla forms.
HunyuanOCR-1.5 is a lightweight end-to-end OCR vision-language model that improves efficiency via DFlash (6.37x inference speedup) and capability via Agentic Data Flow, achieving top-tier performance on document parsing, OCR, and multilingual tasks.
LEDGER is a new benchmark for evaluating long-context capabilities of LLMs on corporate annual reports, providing 4,999 digitized reports with 31 financial KPIs and three evaluation tasks spanning retrieval and extraction.
Jerry Liu's team is presenting ParseBench, a comprehensive document understanding benchmark for VLMs, at CVPR 2026. The benchmark includes 2,000 pages of real-world enterprise documents with evaluation metrics for tables, charts, and visual grounding.
MM-BizRAG is a multimodal retrieval-augmented generation system for enterprise Q&A that uses document structure-aware splitting and layout-aware parsing to outperform vision-centric baselines by up to 32% on heterogeneous enterprise documents. The paper also introduces FastRAGEval, a cost-efficient LLM-based evaluation metric with stronger human alignment than RAGChecker.