Knowledge catalogue
A 36‑page ArXiv whitepaper titled **ExtractBench** was released by Jerry Liu (jerryjliu0), describing a large‑scale, schema‑guided benchmark for real‑world document extraction from complex enterprise
The most dangerous document extraction failure isn't a wrong value. It's a missing row that looks like nothing is wrong. We released ExtractBench yesterday: 370 enterprise docs, 14 systems. The hardes
We wrote a 36-page ArXiv whitepaper on ExtractBench 🧑🔬 , our effort to create the most comprehensive, schema-guided, real-world document extraction benchmark. It’s extremely detailed and covers every
ExtractBench is one of the most comprehensive benchmarks for real-world document extraction. ✅ It covers 4869 pages, across 67 document types, spanning 8 real-world domains: finance, energy, gov, auto
Introducing ExtractBench, the most comprehensive benchmark for information extraction from complex enterprise documents. The latest models are pushing the frontier of coding and knowledge work, but su
Introducing 𝗘𝘅𝘁𝗿𝗮𝗰𝘁𝗕𝗲𝗻𝗰𝗵: the most comprehensive benchmark for information extraction from complex enterprise documents. Our applied research team tested: 14 systems — frontier VLMs, coding agents, ex
A downside with VLM-based parsing is that they’re generally slower than text-based heuristic approaches. As a result they add latency to any ad-hoc file processing *in-the agent loop* (e.g. if you upl
是我见过OCR效果最好的一个了，人眼都难以分辨的，它能处理的十分准确，而且速度非常的快。llamaindex 真不愧是文档解析界的一哥啊。 The best 'raw' frontier model for document parsing is gemini 3 flash, but the issue is that since then the flash models have gotte
Simon @disiok proposed a novel IQ test for language models, asking whether an LLM becomes smarter or dumber after being fine‑tuned on a user’s stream of consciousness. The tweet was posted on 9 Aug 20
The best 'raw' frontier model for document parsing is gemini 3 flash, but the issue is that since then the flash models have gotten 3x more expensive while flatlining on visual recognition across comp
The future of FDE work seems closely related with all work around evals/posttraining/RL envs. FDEs are effectively responsible for the following: 1. Define the business problem. 2. Codify the business
We're not Palantir, but we do think a lot about evals and hillclimbing w.r.t. document processing. If you have really hairy problems around large-scale extraction over complex, real-world document cor
LiteParse can now extract structured data from your PDF in milliseconds: ✅ checkbox states ✅ annotations ✅ vector graphics ✅ word-level bounding boxes It is the most comprehensive, accurate (and fast)
We've built a new feature in LlamaParse that lets you automatically extract any complex form into a structured JSON output 📋🤖 The best part is there's no schema needed! We will systematically detect a
Fully onboard with productizing hillclimbing as an automated service for any agentic task. I've had the fortune of knowing @silennai since the AutoGPT days, and I know that him and Kion are going to d
Document OCR is not Getting Commoditized (by Frontier Models) The most common question I get is whether frontier models are going to eat all document processing solutions - just screenshot the page an
'OCR is just a feature now. Frontier models will eat it.' We hear this constantly. The data says otherwise. Across three GPT generations, parsing accuracy gained ~24 points, while cost per page 4x'd.
The tweet is a repost of Sarah Chieng’s “ggp run” experiment where people recite names of AI‑native companies in alphabetical order, making time feel like it passes faster. The list presented alphabet
Parsing a W-2 into markdown was always the easy part. Getting the fields out was a second pipeline: define a schema, map the fields, handle the edge cases. Set processing_options.forms='𝗲𝗻𝗿𝗶𝗰𝗵', and L
OpenAI’s entire growth loop: new models, price cuts, and Tibo’s token resets major price cuts today: *80% drop for GPT-5.6 Luna, now 0.20 per million input tokens and 1.20 per million output *20% drop
On July 31, 2026 at 01:40 AM, Twitter user Murtaza Khomusi posted a “throwback” tweet reminiscing about opening his first iPhone in 2007 (“2k7 vibez”) and questioned whether this marked the start of t
We created a document OCR router that can estimate the complexity of every single page and parse it with the relevant mode 💫 * Some pages are full of native text, which can be directly handled with Li
We have an idea for dinner 2 already :) But what ideas are you all interested in? - technical topics like rl envs, continual learning, cloud agents, world models - general startup / company building f
Introducing the Parse Gateway There's been an explosion of interest in model routing - you don't always need the best model for every task. That is especially true for document parsing 📄🔀: - Some page
Not every page in your PDF needs the same treatment. A scanned cover, a dense table, a clean text page, a figure-heavy diagram.... most parsing pipelines throw all of them at the same parser, forcing
We’re excited to rollout an official batch parsing experience to LlamaParse. ✅ Instead of hitting our APIs one file at a time, create a batch of 10k at once. ✅ Get a dedicated UI where you can audit t
Yesterday I cohosted a dinner with @dexhorthy with a wonderful group of founders, to talk about agent loops and loop engineering. Some interesting insights: * Most of our group was *not* actively usin
Batch parsing used to mean writing API scripts. Now it's a button. 🦙 Parse or extract across up to 10,000 files in one run, straight from the LlamaParse UI. Point it at a folder and go — no code requi
Sakana is at the frontier of having fun and I respect that We are excited to share our latest work, together with @nyuniversity: 'Dream-Cubed: Controllable Generative Modeling in Minecraft by Training
If I only went off X posts, I'd think Ramp was an AI lab We’re open-sourcing PorTAL, our framework for shared task representations and cross model LoRA adaptation. It now spans from hybrid attention m
This part is spot on: > 'Overall, we found that we were over-constraining Claude Code...while these constraints were once needed to avoid worst case scenarios, we have since found we can delete many o
Title: “don’t do this” refers to a conversation on Twitter where Jerry Liu warns against a particular action, while Julian Schrittwieser comments on a separate thread expressing excitement that Jensen
The Opus 5 system card itself is a fun PDF to parse. It's 193 pages and stacked with labeled and unlabeled charts 📊 LlamaParse does a surprisingly good job on agentic (1.25c per page) and agentic plus
Am I supposed to interpret that it's better than Fable 5 from the benchmarks? Or it's ~close but cheaper? Introducing Claude Opus 5. It's a thoughtful and proactive model that comes close to the front
An interesting thing I'm observing from the blog/system card is that on a good chunk of the reported benchmarks (~20-30% from a skim), Opus 5 max thinking leads to a degradation in performance compare
We benchmarked Opus 5 comprehensively on document understanding through ParseBench. It is roughly on par with Opus 4.8 - it does a few points worse on dense tables, but does slightly better on parsing
LlamaParse can't solve a conjecture, but it can parse this as a PDF 🫡 Dinitz-Garg-Goemans conjecture is false. This graph theory problem was open for ~30 years. The graph below has fractional flow cos
You asked, we delivered! 🚀 Expanding our developer toolkit with: 🐹 Go SDK 🏢 Java SDK >_ CLI Parse a document from Go or Java in a few lines, or straight from your terminal. 🦙 Grab an API key and start
🆕 ActiveGraph: The Log is the Agent my talk from AI Engineer is live!!! 😆 https://www.youtube.com/watch?v=khVX_BUnEwU it's about @activegraphai, an event-sourced graph runtime for building durable lon
Are there MBA programs teaching fear marketing yet We're partnering with @huggingface to investigate an unprecedented security incident. Cyber-capable OpenAI models compromised Hugging Face production
i used to do this a lot. but then i felt like i was getting progressively dumber at writing. i think there's some value in forcing your own brain to sharpen your thinking, instead of offloading it to
Incredible Introducing the world's fastest tokenizer implementation, Gigatoken! Gigatoken is ~500-1000x faster than HuggingFace, and ~100x faster than OpenAI's tiktoken for most tokenizer definitions
One of the hardest parts of document parsing is getting granular attribution and bounding boxes. This lets you ground each piece of text in the specific place in the document it came from. We’ve done
We benchmarked Gemini 3.6 Flash and Gemini 3.5 Flash Lite on document understanding. We compared against their prior versions - Gemini 3.5 Flash and Gemini 3.1 Flash Lite. 1️⃣ Gemini 3.6 Flash has rou
We dropped some new LlamaDrip 🧢 Fear of Docs LlamaParse The team flew into SF for a week onsite. 🌉 2x'd in size since we last did this — first time this many of us have been in the same room. The reca
We're parsing some of the hardest financial documents into clean, plaintext/structured outputs through a live webinar. Come check it out! https://watch.getcontrast.io/register/llamaindex-from-complex-
Your agentic workflows are only as good as the context you feed them. In finance, that context is stuck inside your messiest documents: dense tables, footnoted adjustments, and the details buried in t
Can’t tell if the PR reads more like a security incident or a product release… We suspected last week's cyberattack might have come from a frontier lab, given the sophistication of the agent. Turns ou
feeling nostalgic, my favorite blogs on RL & reward hacking https://www.alexirpan.com/2018/02/14/rl-hard.html https://lilianweng.github.io/posts/2024-11-28-reward-hacking/ TLDR: An openai model, durin
The team flew into SF for a week onsite. 🌉 2x'd in size since we last did this — first time this many of us have been in the same room. The recap: 🧠 Locked in our roadmap for AI agent document infra 🎨
Extracting information from millions of documents at scale used to take an insane number of human hours. Even with recent OCR + document AI tech, humans would still have to spend a lot of time careful
We're part of the Amazon Web Services (AWS) AI Builder Lab in New York on Friday, July 24 - a Clash of Agents competition with OpenAI, LangChain, HiddenLayer, Protopia AI, Fiddler AI, and Coder. One d
Agreed with this. people underestimate the importance of good abstractions and maintainability. part of the reason LLMs/AI are so popular in the first place is because of the ease of using the models
Your extraction schema is now a conversation away. 📄 Upload a doc — the agent drafts the schema from it 💬 Want changes? Just ask 📚 Or grab a template and go Writing JSON Schema by hand? That's
congrats on the launch! As the task horizons on agents gets longer, we need more work evaluating and training models to be better in long-running, open-ended, evolving real-world environments. Today w
Ever wanted to quickly turn a PDF into clean text to paste into your favorite AI agent, without having use CLIs or open the browser? We built exactly that. Using @TauriAp ps, with a Rust backend power
Parse any PDF into agent-ready context. Fully local, free desktop app. No model / internet connection required! Ever wanted to quickly turn a PDF into clean text to paste into your favorite AI agent,
There is a massive opportunity for any startup to build specialized, domain-specific workflows that are at the Pareto frontier in accuracy, cost, and latency. It is clear that not every task requires
Today we present Morpheus, a persistent enterprise simulation platform designed to make Continual Learning a reality. Morpheus is the world’s first real world Reinforcement Learning environment. Every
Actually I did a quick test. I like the ChatGPT Work/Codex split better than Claude Cowork/Code. The interface is much more unified. The functionality is effectively the same. The chat history is shar