LlamaIndex 🦙

What happens when a table contains tables?

Usually, a mess. With LlamaParse, it comes out clean.

Doc of the Week: Micron’s latest earnings deck. Even with two business units and same row names, LlamaParse keeps all 18 values under the right header.

Try it out on your own earnings deck: llamaindex.ai/llamaparse

Markdown is all you need.

(Mostly.)

A parser can get every word on the page right and still lose which column a number belongs to. Then your model has to guess.

Markdown keeps headings, lists, and tables intact, and it stays readable when you're debugging a bad answer. For tables with merged headers, we switch to HTML.

Read our breakdown on why it's our default output for parsing below! ⬇️

Today we're announcing OpenDocRouter: every model for document parsing under one API.

There are ~4,000 OCR models on Hugging Face, and the frontier labs ship a new one nearly every month. Whatever's best for your docs today won't be by Q1.

So stop asking which model to use for doc parsing. Ask how fast you can switch.

✅ Switch models in one line: same request, same markdown output, no new prompts or integrations
✅ Any model you want: 10 frontier and open-source models at launch, including Claude Opus 5.5, Gemini 3.8 Flash, GPT-6 Luna, MinerU2.5-Pro and PaddleOCR-VL-1.6
✅ Choose with receipts: every model scored on ParseBench for quality and cost
✅ Traceable output from any model: "layout: true" adds grounded bounding boxes and the same layout classes, even for models that don't support it natively
✅ Pay only for what works: per-token pricing, failed pages never charged, top up from $25

The spread is the point: $0.86 to $48.82 per 1,000 pages, depending on what your documents actually need.

Live now → opendocrouter.ai

The biggest night of SF Tech Week is black tie.

120+ agent founders, 50 investors, and live jazz at the Redwood Room on Friday for the Agent Economy Gala. Invite only.

We're excited to be partnering with @SpaceX, @Daytonaio, @lemma_, @Tiny_Fish, @thelinqapp, @Obvious & more. Link to request an invite in below or comment what you're building!

OCR is dead 🪦 Long live agentic OCR!

Traditional OCR makes one pass and hands back whatever text it got. Tables get flattened, charts disappear, and multi-column layouts come out scrambled. Nothing checks the output.

Agentic OCR treats parsing as a loop instead:

✅ layout-aware reading order
✅ smart routing of hard elements to the right model
✅ multi-pass verification and self-correction
✅ multimodal parsing of charts, images, and complex tables

Logan Markewich, LlamaIndex's Head of Open Source, wrote about what that shift looks like in practice and where it's still struggling. Link to the breakdown in the comments below.

We got extremely good at extracting a scanned from layer by layer 📋 - from printed fields, overlaid text, handwriting, and reviewer annotations.

For instance, on an annotated permit, our Extract agents can extract out both the original permit number but also any drawn/handwritten annotations on top. they're also grounded exactly in the source text.

our agents reduce error rates by 2x+, at a small fraction of the price compared to using a frontier model.

Our Extract v2.5 agents are SOTA in price/performance, and scanned forms are just one example. Come check it out!

Extract v2.5: llamaindex.ai/blog/introduci…

Available in LlamaParse: cloud.llamaindex.ai

The BEA's new Q2 GDP-by-state was released, but the key data is entirely in an infographic, without a text-layer.

LlamaParse's agentic parsing tier understands documents like this by default, and faithfully records data, turning the page into a 52-row table.

Source: bea.gov/sites/default/…

How many people show up on a Tuesday night to talk about document processing for AI agents? 🤔

In New York, enough to fill the room and start a waitlist. In San Francisco, almost 600.

We were in both cities on the same night to talk about one problem: agents don't read documents the way people do. Look at an invoice and you can tell the monthly rate from the total in a second. An agent often gets the words without the layout, so it has to guess.

In New York, we partnered with @usekernel on @braintrust's Agent Builders Night. In San Francisco, we partnered with @WorkOS on @daytonaio's AI Builders event. Thanks to them for having us, and to Abrar and Yong from our team for speaking!

Document parsing requires a lot of on-the-fly decision making.

We explored the early approaches to Jev and Jev-like models on several documents tasks like orientation detection, language detection, routing, and more.
youtube.com/watch?v=MLZ1dU…

Parsing forms still trips up the latest frontier VLMs.

A form isn't text on a page. It's a set of fields, grouped into sections, each tied to a specific box. That's why forms need purpose-built parsing, not a bigger general model:

✅️ Detect every field, not just the obvious ones
✅️ Keep the hierarchy of sections and fields
✅️ Tie every value to the exact box it came from
✅️ Understand handwriting and checkmarks

Our latest blogpost breaks down where VLMs fail on real W-2s, 1040s, W-9s and scanned W-4s, and a custom cookbook for LlamaParse to handle them at a fraction of the cost 👇

llamaindex.ai/blog/why-vlms-…

A blank cell can change the meaning of a forecast.

Four times a year, the Fed's 18 top policymakers each put their forecasts on paper: where growth, jobs, inflation, and interest rates are headed. This is September 2026's edition, home of the "dot plot" that markets treat as the Fed tipping its hand. This release is the closest thing to the Fed saying what it plans to do. Most analysts will want to throw this documents to an AI agent, but this messy doc is dense: full of complex tables and charts that hold valuable context. All things that frequently trip up raw LLM APIs.

The Fed’s September 2026 projections table includes a 2029 column, but its June comparison row leaves that cell empty. This is a common but silent failure point for document parsers.

We parsed page 2 with LlamaParse and checked the displayed GDP median excerpt against the original PDF. All nine numbers matched and the data stays aligned in the returned HTML.

Try LlamaParse on a table where headers and missing cells matter. cloud.llamaindex.ai/signup

Source: federalreserve.gov/monetarypolicy…

confidence scores only matter if they help you decide what to automate.

for document extraction, that usually means knowing how much work you can safely accept at a given precision target. in our latest post, we look at confidence scoring through that lens, including:

✅️ confidence cutoffs
✅️ precision vs. recall
✅️ score coverage
✅️ score granularity
✅️ human review volume

using ExtractBench, we compare how different extraction systems perform after confidence filtering. at a 97% precision target, LlamaParse Agentic Plus reached 66.48% recall on expected fields after filtering.

the useful part of a confidence score isn’t the number itself. it’s whether you can use it to control automation and review in production.

👉️ read the full post: llamaindex.ai/blog/what-make…

The fastest PDF-to-Markdown parser just got faster. ⚡️

With LiteParse v2.14.6, text-based PDFs parse about 25% faster.

Across realistic documents, LiteParse processed pages at 2.8ms/page and 1.5× faster than the next-fastest local parser.

LiteParse is open source and runs locally in Python, Node.js, Rust, or directly in the browser.

Grab v2.14.6 → github.com/run-llama/lite…
Bench Docs → huggingface.co/datasets/llama…

Grounded Confidence is here for Extract! 🦙

When your agents and workflows depend on extracted data, you need to know how accurate that data is.

We’ve added confidence scores to give you a better read on extraction accuracy, field by field. Use them to decide which results your app can accept automatically and which need human review.

Available on Cost Effective, Agentic, and Agentic Plus.

Try it out on your docs -> developers.llamaindex.ai/llamaparse/ext…

Introducing the first in our Parsed by LlamaParse series. We're outlining the stakes of 'getting parsing wrong' in consequential documents.

We’re starting with the U.S. Energy Information Administration’s September 2026 Short-Term Energy Outlook, a dense government report covering energy supply, demand, prices, and forecasts.

Table 7a alone packs multiple years, quarters, row hierarchies, units, and footnotes into one electricity-industry table.

Take 1,186.

Parsed correctly, it means electricity sales to ultimate customers in Q3 2026, measured in billion kilowatthours.

Parsed incorrectly, it could be assigned to the wrong quarter, metric, or unit , which means the error can flow straight into a dashboard, forecast, alert, or AI application.

And the values aren’t the only thing that matters. Footnotes define the data too: “small-scale solar,” for example, refers to systems under one megawatt, not solar generation overall.

LlamaParse preserves the structure and context that make document data usable downstream: whether you’re populating a database, updating a dashboard, running forecasting workflows, or building an AI app over complex documents.

Source doc here: eia.gov/outlooks/steo/

It was great to be at Connected Stack last week with founders and builders working on what’s next in enterprise AI.

@jerryjliu0 joined the Founder Flash Talks to talk about a problem every enterprise agent eventually runs into: messy, complex documents that general-purpose models struggle to read.

Agents are the new knowledge workers, and we are building the document infrastructure for agents.

Thanks @trueventures and @GreylockVC for having us! 📸

Ready to learn how to build document processing pipelines for insurance agents?

Join LlamaIndex Solutions Architect Abrar Mahi for a webinar on turning insurance documents into structured data for underwriting, policy review, and claims workflows.

Using LlamaParse and Extract, we’ll cover:

✅️ Parsing complex accord forms, policy documents and other popular document types in the insurance industry.
✅️ Extracting policy details, property information, and claims history into a defined schema.
✅️ Verifying extracted values with citations and bounding boxes.
✅️ Combining confidence scores with validation rules to decide what proceeds automatically and what needs human review.

We’ll bring these steps together in a working pipeline you can adapt to your own workflows.

Register: watch.getcontrast.io/register/llama…

keeping our SDKs up to date can be a pain. every API change brings another round of updates, tests, and releases. and each SDK should still feel like it was written by someone who actually uses the language.

Stainless helped us do that for LlamaParse. It also pushed us to improve the API itself. inconsistent names and schemas become harder to ignore when they show up in the code developers use.

with the Stainless team joining Anthropic, George He and Yong Park wrote about what worked, what we learned, and why changing SDK generators takes more care than you might expect. -> llamaindex.ai/blog/au-revoir…

thanks to the Stainless team for saving us a lot of SDK work. 👋

au revoir, Stainless

just-in-time OCR is all the rage.

most pipelines parse every page before anyone asks a question. for an agent working through an ad-hoc data room, that's slow, expensive, and most of those pages never get read.

the better pattern is just-in-time OCR in two passes:

✅️ LiteParse (free, OSS, Rust, 50+ formats) does a fast layout-aware first pass: spatial text, bounding boxes, headings, tables, and a per-page complexity flag. a full data room in 32 seconds.

✅️ LlamaParse zooms in on only the pages that need it, by page number, and returns cell-level tables, bounding boxes, and confidence scores. the rest fills in the background.

pypdf and pdftotext can't do the first pass well. parsing everything up front can't do it cheaply. two passes gets you both.

full breakdown with numbers: llamaindex.ai/blog/just-in-t…

confidence scores in LlamaParse just got an upgrade 🦸‍♀️

our new high-effort mode provides granular page-level scores with text explanations to help you intimately understand parsing quality of your documents. we even refer back to the original document for an extra check when generating the score.

use high-effort only when you need it, at 5 additional credits per page.

try it on your docs: developers.llamaindex.ai/llamaparse/par…