Apple dropped new model LensVLM-9B on Hugging Face and it Beats RAG-style baselines up…
This is a AI post classified by Jev as RAG & memory (a model release), kept by the AI Radar because it carries real work, not commentary.
Apple dropped new model LensVLM-9B on Hugging Face and it Beats RAG-style baselines up to 10.1x. It’s a 9B Qwen3.5 finetune that turns long documents into compressed page images (5-15x), scans them visually, then expands only the relevant pages back into full text. - Render the document as tiny page images - Scan the compressed pages - Expand only the pages that actually answer the question - Near full-text accuracy at 4.3x compression. Result: token-efficient long-context QA with near full-text accuracy at 4.3x effective compression. Paper says it beats RAG and other baselines up to 10.1x
Posted by Md Ismail Šojal 🕷️ (56.1k followers) 1 h ago · 5 likes · 664 views · view the original post on X. Kept by the AI Radar as RAG & memory. Tools mentioned: lensvlm-9b.
More AI work like this
- we've built the world's most advanced engine for document extraction over complex… — @jerryjliu0
- 🔖 Cookbook Highlight: Build a Grounded Chat Agent with r-1 — @reductoai
- Build an assistant that can store and search what it sees, hears, and is told, all… — @DeepLearningAI
- Digital PDFs or warped phone photos, TeleOCR parses them with one lightweight 1.2B… — @ModelScope2022
- Interesting paper on agent memory stored as a linked markdown wiki. — @omarsar0
- D-RAC: Universal Retrieval-Aware Ingestion — @HuggingPapers
- liteparse is the fastest pdf parser on the planet — @jerryjliu0
- DocJev is the fastest way to classify and split complex document packets ⚡️. (and the… — @jerryjliu0
Every post is read and classified by Jev (TypeSafe): what it is, which market it belongs to, and whether the link is a real tool. 45.7k posts from 5k X accounts over the last 14 days, 1.9k tools, 19 markets. Collected every 5 minutes, fully re-ranked every hour — last update 2026-09-24 02:07 UTC. Full method.