Send me the page that breaks your parser.
Merged-cell tables, 30 year old scans, handwritten claims forms, engineering drawings you name it, we want the worst of the worst and the ones that cause the most pain today.
I'll run it through LlamaParse and post the output. No pitch, no form. Just reply or DM.
Records that span multiple pages: 85.5% → 96.5%.
That's where we most often see customers get stuck with existing tools. With citations on every value, the output is easy to trust.
Great work from the team!
Today we’re introducing Extract v2.5 - a series of frontier agents tuned for document extraction.
The agents (cost-effective, agentic, agentic plus) are tuned for value accuracy and grounding. Our extraction agents outperform Opus 5.5 and GPT-6 Sol while being 30%-4x cheaper.
We’ve made massive improvements on complex extraction over
* long lists (86.1% -> 95.5% on our agentic tier)
* records spanning pages (85.5% -> 96.5%)
* scanned forms (90.9% -> 95.7%)
We’ve also launched the following features:
✅ Advanced citations: we locate bounding boxes for all supporting values for an inferred field, even if there's not an exact match.
✅ Structural Reasoning: we tailor document extraction algorithms depending on the type, layout, and information
Our extraction agents are SOTA in price-performance on ExtractBench, across a wide range of cost points. We are the best tool for document extraction across documents of any complexity.
Blog: https://t.co/TTlTdECU03
All of these are available on LlamaParse: https://t.co/XYZmx5TFz8 . Come check it out!
@alexeheath With chrome you're locked into Gemini. Comet allows you to use models from all top labs directly in your browser. Being able to use the best model for your specific task is a superior experience.
@MostlyMonkey Teaching multiple ppl the best way to do your old job is the best way to scale your contribution. This sounds like an inability to say no to meetings and an inability to coach subordinates.
Also hiring.
"There's this thinking that any full-stack engineer can be an ML engineer with ChatGPT APIs," says @Sprig founder/CEO @ryanglasgow.
But he makes the case for why you might actually need a dedicated ML engineer on your eng team.
🚀 Today, I’m excited to announce that @Sprig is launching an industry-first capability: AI Analysis for Surveys!
⬇️ @VentureBeat covered the exciting news:
https://t.co/Eto0MrA8MH
✨ In addition, we’re also expanding our Free and Starter plans and publicly announcing new marquee customers: PayPal, Figma, Ramp, Peloton Interactive, and Mixpanel.