Click any tag below to further narrow down your results
Links
Indra OCR is a lightweight 1M-parameter model that converts PDFs and images to markdown, text, or JSON at 49 pages per second on CPU—roughly 12-24x faster than Tesseract and PaddleOCR. It handles layout detection, bounding boxes, and can read digital PDFs' text layers while OCRing only missing content.
- Speed and size: Processes 49 pages/second on CPU with ~230ms latency per page, weighs just 18MB, and uses ONNX Runtime instead of PyTorch
- Accuracy beats competitors: Scores 41.3 on OmniParseBench text (vs. PaddleOCR's 35.8 and Tesseract's 19.3) and 62.1 on headline score (vs. PaddleOCR's 58.4)
- Layout intelligence: Detects and labels blocks (headings, tables, lists, captions, footnotes) with bounding boxes and reading order, plus optional LaTeX math recognition
- Built for print documents: Not suitable for handwriting, complex tables, degraded scans, or scripts outside Latin/Cyrillic/Greek/Devanagari/Bengali/Chinese; Datalab offers Chandra model for these cases
Chandra OCR 2 is Datalab's optical character recognition model that converts documents to markdown, HTML, or JSON while preserving layout. It scores 85.8% on the olmOCR benchmark and supports 90+ languages with strong performance on tables, math, handwriting, and complex layouts.
- Chandra 2 achieves 85.8% on olmOCR benchmark (state-of-art), a 12% improvement over Chandra 1's 77.8% multilingual score, with particular strength in math, tables, and complex layouts
- Available through multiple deployment methods: HuggingFace Transformers, vLLM, SGLang, and Docker, with quickstart CLI tools for local use
- Handles 90+ languages, extracts images with captions, reconstructs forms accurately including checkboxes, and supports handwriting recognition across diverse document types
LiteParse is an open-source PDF parser that extracts text with spatial positioning and bounding boxes entirely on your machine, running 2-5ms per page without cloud dependencies or LLMs. It handles PDFs, Word docs, spreadsheets, and images, with optional OCR for scanned content.
- Benchmarks show LiteParse outperforms free competitors (pdf-inspector, opendataloader, markitdown) across three major document parsing tests, scoring 0.389 on ParseBench vs 0.283 for the next best model-free tool.
- Speed comes from avoiding ML models — uses PDFium for text extraction, optional Tesseract or PaddleOCR for scanned pages, and spatial layout reconstruction via grid projection.
- Available in Python, Node.js, Rust, and browser (WASM) with CLI tools; includes worker pool mode for parallel parsing with per-document timeouts to prevent pipeline stalls.
Xberg is a single engine for detecting, reading, OCR’ing, and extracting text, tables, metadata, and structured data from over 100 document formats and 115 file extensions. It offers transcription, embeddings, layout reconstruction, schema-driven JSON extraction, and code intelligence via language bindings in Rust, Python, Go, Java, and more. You can run it as a library, CLI, REST API, or MCP server with configurable Cargo features.
- One Rust-core engine handles 101 document formats (115 extensions) plus 371 programming languages for code intelligence
- Ships as 15 language bindings (Python, Go, Java, Ruby, PHP, Elixir, C#, TypeScript, etc.) plus CLI, REST API, and MCP server
- Combines OCR (Tesseract/PaddleOCR/VLM), layout reconstruction (PP-DocLayout-V3, RT-DETR), and table extraction (TATR, SLANet) with schema-driven JSON output via local or hosted LLMs
- Runs GPU-free with multi-GB streaming support and safeguards against zip bombs and excessive nesting/compression
Mistral OCR 4 extracts text from PDFs, DOCs and more while also returning bounding boxes, block types and per-word confidence. It supports 170 languages, runs in a single container for self-hosted deployments, and outperforms rivals on human and automated benchmarks.
- Mistral OCR 4 doesn't just extract text—it returns bounding boxes, block types (titles, tables, equations, signatures), and per-word confidence scores, and can be self-hosted in a single container for data-sovereignty needs
- Beat every tested competitor in human evaluations across 600+ documents/12+ languages with a 72% average win rate, and scored 85.20 on OlmOCRBench and 93.07 on OmniDocBench
- Roughly 4x faster than some enterprise OCR providers, and 8x cheaper with 17x lower latency than leading agentic document parsers on finance datasets with charts/figures
- Priced at $4 per 1,000 pages via API ($2 with batch discount) or $5 per 1,000 pages through the no-code Document AI interface
Unlimited-OCR is an open-source OCR framework from Baidu that extends Deepseek-OCR to handle single images, multi-page documents, and PDFs. It provides Hugging Face transformer models, PyMuPDF conversion, and an OpenAI-compatible SGLang server for streaming or batch inference. The repo includes setup instructions, example scripts, and configuration for GPU-accelerated parsing.
- Extends Deepseek-OCR to handle not just single images but multi-page documents and PDFs, with a 32,768-token context window for long-horizon parsing
- Includes an OpenAI-compatible SGLang server supporting streaming inference, plus a batch script that auto-launches the server and processes image/PDF folders concurrently
- PDF handling pipeline renders pages at 300 DPI via PyMuPDF, then feeds them through model.infer_multi with a no-repeat-ngram constraint to avoid repetitive output
- Offers two preset image configs ("gundam" with crop mode at 1024/640, "base" at fixed 1024x1024) for trading off resolution and processing mode
This daily roundup highlights new AI tools, models, and research—from Mistral’s OCR 4 and ByteDance’s Seedance 2.5 video generator to Anthropic’s Claude Tag and IBM’s CUGA agent harness. It also covers security deep dives on indirect prompt injection, industry moves like OpenAI’s bidirectional voice and US‐Meta AI reviews, plus several open-source releases.
- Mistral OCR 4 handles 170 languages and runs 4x faster than rivals, especially on low-resource scripts
- ByteDance's Seedance 2.5 generates 30-second 4K videos from a prompt plus up to 50 reference images, clips, or audio files
- Airbyte Agents cuts tool calls by 40%, token use by 80%, and multi-source query costs by 90% by indexing business data instead of querying APIs live
- OpenAI's new Bidirectional Voice Mode can hold real-time back-and-forth conversations and even sing or beatbox within copyright limits
Datalab’s 4 billion-parameter Chandra OCR 2 outperforms GPT-4o and Gemini across independent and multilingual benchmarks, handling complex layouts, math notation, flowcharts and 90 languages with state-of-the-art accuracy. It’s available under Apache 2.0 code with a modified OpenRAIL-M license for weights, runs locally via HuggingFace or vLLM, and doubles throughput over its predecessor.
- Chandra OCR 2 (4B params, open-weight) scored 85.9% on olmOCR vs GPT-4o's 69.9%, and beat Gemini/GPT-5 Mini on multilingual benchmarks, with huge gains on South Asian scripts (Kannada +42.6, Malayalam +46.2, Telugu +39.1).
- It processes full pages in one pass rather than splitting into blocks, giving it an edge on tables, nested headers, checkboxes, handwritten math, and flowcharts exported as Mermaid diagrams.
- Despite shrinking from 9B to 4B parameters, throughput doubled to ~2 pages/sec on an H100 while accuracy improved.
- Code is Apache 2.0 and installable via pip/Docker, but weights use a modified OpenRAIL-M license requiring a paid commercial license for larger companies.
Chandra OCR 2, a 4 billion-parameter model from Datalab, outperforms GPT-4o and Gemini on AllenAI’s olmOCR benchmark and a 90-language test while halving the model size. It preserves layout, reads complex tables and math notation, converts diagrams to Mermaid, and runs at two pages per second on an NVIDIA H100. The code is Apache 2.0 but the model weights use an OpenRAIL-M license with commercial restrictions.
- Chandra OCR 2 scores 85.9% on olmOCR vs GPT-4o's 69.9%, while cutting model size from 9B to 4B parameters and hitting ~2 pages/sec on an H100
- Multilingual performance beats Gemini 2.5 Flash and GPT-5 Mini, with 40-46 point gains on scripts like Kannada, Malayalam and Telugu over Chandra 1
- Weights carry an OpenRAIL-M license requiring a paid commercial license once a company exceeds $2M in funding or revenue, despite Apache 2.0 code
- Handwriting recognition remains weak, dropping to ~50.4% accuracy on complex forms despite strong printed-text and table/math handling
The article features a discussion page for the dataset "100-handwritten-medical-records" hosted on Hugging Face. It provides an overview of the dataset, including its modalities, formats, size, and community interactions such as discussions and pull requests. Users can engage with the dataset and contribute to its development.
- The dataset contains 100 handwritten medical records hosted on Hugging Face
- It's set up as a community discussion hub, allowing users to contribute via discussions and pull requests
- No actual discussions or contributions appear to have been posted yet based on the available content