Baptiste Aubertin retweeted
The anomaly of the LightOn R&D alumnis need to be studied honestly
@iacopo_poli really created something special
IFYKYK
Baptiste Aubertin retweeted
Today we’re releasing LightOnOCR-3, the third generation of our OCR models—a joint effort with @staghado and @baptaubertin!
Here’s how we built its grounding data, and how intermediate checkpoints helped improve the annotations used for subsequent training. 🧵
Baptiste Aubertin retweeted
🚀 LightOnOCR-3 is out! 🦉
This time we go beyond OCR: text transcription, document layout, image descriptions and chart extraction, all in one model.
We release 3 sizes: 0.8B, 1B and 4B, all under Apache 2.0!
Everything is open, including the scripts to reproduce our numbers. Give it a try and tell us what breaks.
Blog: huggingface.co/blog/lightona…
Demo: huggingface.co/spaces/lighto…
Code: github.com/lightonai/LightOn…
Huge thanks to my colleagues @staghado and @CavaillesAdrien 🚀, and to the whole @LightOnIO team for making this happen 🙏
What’s new: add the word “grounding” to your prompt and every block comes back with a labeled box. Images get a short description, and charts come back as data tables.
Without it, you get the same clean Markdown as before.
🚀 We just updated LightOnOCR! 🦉
LightOnOCR-3 is faster, more accurate, and brings new capabilities: it now tells you where everything is on the page, describes images, and pulls the numbers out of charts.
Three sizes (0.8B, 1B, 4B), still Apache 2.0.
Baptiste Aubertin retweeted
World-leading document recognition. Beyond text extraction.
Introducing LightOnOCR-3, our new family of state-of-the-art AI models for document intelligence.
Building on LightOnOCR-2, it recognizes text and handwriting, describes images, extracts data from charts and preserves document structure, all in a single processing step.
🏆 Leading benchmark results : ahead of Chandra-OCR-2, Mistral OCR 4.1 and dots.mocr on OlmOCR-Bench, and first among open-weight models on ParseBench.
📊 More information from every document : turn text, tables, images and charts into content your business applications can use.
⚡Compact and efficient: 0.8B and 4B versions, with document processing up to twice as fast as LightOnOCR-2.
🔓Apache 2.0 license: use, adapt and deploy the models on your own infrastructure.
From handwritten administrative files to financial reports and technical documents, make more of your information searchable and usable.
👉 Explore the models and benchmark results: huggingface.co/blog/lightona…
👉🏻Explore the model page on our Website : lighton.ai/research/lightono…
#LightOn #DocumentIntelligence #OCR
Baptiste Aubertin retweeted
A few months ago @treygrainger and @softwaredoug invited me to guest lecture in the AI-Powered Search course!
I have been a fan of the book for a long time, so I was beyond hyped and went deep in the multi-vector rabbit hole. The result is my 64-slide monster baby 😂
Baptiste Aubertin retweeted
LightOn Search reaches 5th place on BrowseComp-Plus with a strong accuracy-per-search-call ratio
LightOn Search is our search API, powered by our best models 😁 @LightOnIO
Baptiste Aubertin retweeted
Today, we're introducing LightOn Console.
⚙️ Three endpoints:
/Parse any documents
/Extract structured data
/Search enterprise knowledge with citations
🔌 Built-in connectors. MCP-ready. Governance enforced at the chunk level.
No infrastructure. No pipeline maintenance. No dedicated retrieval team required.
Make your enterprise knowledge agent-readable now!
Read the launch announcement: lighton.ai/lighton-blogs/int…
Test it now: console.lighton.ai/
Baptiste Aubertin retweeted
Baptiste Aubertin retweeted
Releasing ColGREP and LateOn-Code models 🚀
ColGREP is a multi-vector search tool built in Rust made for coding agents. It's an hybrid grep which supports both grep features and semantic retrieval. Run 100% locally.
You get two SOTA code retrieval model within ColGREP
Baptiste Aubertin retweeted
Introducing Better Live Text : Modern OCR for macOS ✨
Very useful for grabbing equations in perfect LaTeX when reading papers!
Baptiste Aubertin retweeted
The LightOnOCR preprint is now on arXiv, if you want to learn more about how we built the model from data curation, pretraining to RL, give it a read!
arxiv.org/abs/2601.14251
Baptiste Aubertin retweeted
Lighton released the new sota 1B OCR model with A2.0 license 🔥
> super fast & cheap ⚡️ 5× faster than dots.ocr, 2× faster than PaddleOCR-VL and costs <$0.01 per 1k pages
> detects images, handles various media and layouts
> comes with transformers support day-0! 🤗
Baptiste Aubertin retweeted
🚀 LightOnOCR-2-1B 🦉 is out, a major update to LightOnOCR.
1B parameters, end-to-end multilingual OCR, and it beats models 9× larger on OlmOCR-Bench while being much faster.
PDF/page in, clean ordered Markdown out, with optional image localization (bbox variants).
Baptiste Aubertin retweeted
A Halloween gift 🎃
New finetuning notebook for LightOnOCR-1B:
• Supports both Full and LoRA training
• Supports FineVision 🤗 subsets incl. OlmOCR-mix & handwritten IAM
• ~12 min/epoch on one H100 (can also runs on Colab!)
Baptiste Aubertin retweeted
Looks like @LightOnIO's LightOnOCR is in good company.
Finetunable, state of the art and cheap!
Awesome job by @staghado , @CavaillesAdrien , @baptaubertin and the team.
In-Depth 👇 lighton.ai/lighton-blogs/mak…
