[metadata]
app-built-at: 2026-11-06T19:25:30.065Z
app-version: 0.0.1+production#d5d1bea
description: Datalab converts PDFs, scans, spreadsheets, and slides into Markdown, HTML, JSON, or structured fields, with a free monthly allowance and pricing per 1,000 pages.
og:description: Datalab converts PDFs, scans, spreadsheets, and slides into Markdown, HTML, JSON, or structured fields, with a free monthly allowance and pricing per 1,000 pages.
og:image: https://www.datalab.to/og-default.png
og:site_name: Datalab
og:title: High-Precision Document Parsing and Extraction API
og:type: website
og:url: https://www.datalab.to/
posthog-host: https://p.datalab.to
posthog-key: phc_tdzhvtMRrjRsRbBNByNWL1Enk1Vym1zOCNrbZ45R2QW
twitter:card: summary_large_image
twitter:description: Datalab converts PDFs, scans, spreadsheets, and slides into Markdown, HTML, JSON, or structured fields, with a free monthly allowance and pricing per 1,000 pages.
twitter:image: https://www.datalab.to/og-default.png
twitter:site: @datalabto
twitter:title: High-Precision Document Parsing and Extraction API
viewport: width=device-width, initial-scale=1

[canonical-links]
https://www.datalab.to/

[document-links]
/
About: /about
Agentic workflows: /use-cases/agentic-workflows
All benchmarks →: /benchmarks
All customer stories →: /customers
All posts →: /blog
Audit & compliance: /use-cases/compliance
Benchmarks · Nov 4, 2026 Which document-parsing benchmarks can you trust? We checked 15 document-parsing benchmarks: we ran seven open models through the 13 with public data, had three judge models count their real errors, and checked which benchmarks rank parsers the way their errors do. Read →: /blog/which-parsing-benchmarks-to-trust
Benchmarks: /benchmarks
Blog: /blog
CASE STUDY →: /blog/datalab-gamma-case-study
CASE STUDY →: /blog/datalab-purchaser-case-study
CASE STUDY →: /blog/datalab-radical-ai-case-study
CASE STUDY →: /blog/datalab-revisiondojo-case-study
Careers: /careers
Chandra Our flagship open OCR model. Works on the messiest docs and the hardest edge cases. 12.6k ★: https://github.com/datalab-to/chandra
Contact: /contact
Customers: /customers
Documentation: https://documentation.datalab.to/
End User License Agreement: /terms/end-user-terms
GH: https://github.com/datalab-to
HF: https://huggingface.co/datalab-to
Indra A 1M-parameter OCR model that reads handwriting, 80+ languages and pages at any angle, fast on a plain CPU.: https://github.com/datalab-to/indra
LI: https://www.linkedin.com/company/datalabto
Log in: /auth/sign_in
Marker + Surya High-throughput OCR engine for resource-constrained environments. Covers 90+ languages. 62k ★: https://github.com/datalab-to
Model training data: /use-cases/model-training
Multi-column OmniParseBench · multi_column tag 94.9% Datalab accurate Datalab accurate 94.9% Datalab balanced 92.7% LlamaParse 91.4% #1 of 10 parsers →: /benchmarks/multi-column
Open source: https://github.com/datalab-to
Overall OmniParseBench 94.1% Datalab accurate Datalab accurate 94.1% Datalab balanced 92.6% Reducto 90.8% +3.3 vs the next vendor API →: /benchmarks/overall
P · 01 Convert Use Chandra 2.5 to convert PDFs, spreadsheets, slides and more to markdown, HTML, and JSON, in fast, balanced, or accurate mode. Add word-level bounding boxes, chart data, and structured figure descriptions. →: /platform/processors/convert
P · 02 Extract Pull the fields you define into JSON, with citations and a confidence score on every field. Four modes, from turbo to accurate. →: /platform/processors/extract
P · 03 Segment Identify natural boundaries between documents that were originally combined into one file. →: /platform/processors/segment
P · 04 Eval Evaluate the quality of output, using our criteria or your own rubrics for quality. →: /platform/processors/eval
Platform overview: /platform
Playground: /playground
Pricing: /pricing
Privacy: /terms/privacy
Processors: /platform/processors
Product Updates · Nov 6, 2026 API submits and result checks, 2x faster Submitting a document now takes 314 ms at the median, down from 670 ms, and result checks are twice as fast. SDK 0.9.1 picks up finished results sooner. Read →: /blog/faster-api-responses
Public Playground: /playground
RAG indexing: /use-cases/rag
Security: https://endless-labs.secureframetrust.com/
See AI Research Labs →: /industries/ai-research-labs
See every processor and its price: /pricing
Sign Up: /auth/sign_up
Sign up: /auth/sign_up
Status: https://status.datalab.to/
Tables OmniParseBench · table tag 95.6% Datalab accurate Datalab accurate 95.6% Extend 94.7% LlamaParse 93.7% #1 of 10 parsers →: /benchmarks/tables
Talk to sales →: /contact
Terms of Use: /terms/terms-of-use
Try Playground →: /playground
Try playground for free →: /playground
Tutorials · Nov 2, 2026 Data extraction vs parsing: Which to use? Choose parsing for document context and extraction for specific fields. Extract, validate, and evaluate financial data with a JSON schema and the Datalab API. Read →: /blog/data-extraction-vs-parsing
We're hiring: /careers
Web-ready PDFs: /use-cases/web-ready-pdfs
X: https://x.com/datalabto
datalab.to/agent-setup.md: https://www.datalab.to/agent-setup.md
olmOCR-bench olmOCR-bench · Chandra 2.5 86.9% Chandra 2.5 Chandra 2.5 86.9% Chandra OCR 2 85.8% LightOnOCR-3-4B 85.4% +1.1 vs Chandra OCR 2 →: /benchmarks/olmocr-bench

[structured-data]
{"@context":"https://schema.org","@type":"Organization","description":"High-precision document intelligence. Transform unstructured content into precise, production-ready data.","logo":"https://www.datalab.to/images/datalab-icon.svg","name":"Datalab","sameAs":["https://twitter.com/datalabto","https://github.com/datalab-to","https://linkedin.com/company/datalabto","https://discord.com/invite/KuZwXNGnfH"],"url":"https://www.datalab.to"}
{"@context":"https://schema.org","@type":"SoftwareApplication","applicationCategory":"DeveloperApplication","description":"Document parsing and extraction API. Converts PDFs, scans, spreadsheets, and slides into markdown, HTML, JSON, or structured fields.","name":"Datalab","offers":{"@type":"Offer","description":"Free monthly usage allowance, then usage-based pricing per 1,000 pages. See https://www.datalab.to/pricing.md.","price":"0","priceCurrency":"USD"},"operatingSystem":"Web, Linux","url":"https://www.datalab.to"}

[page-chrome]
We're hiring
Playground
Product
▾
Solutions
▾
Customers
Pricing
Resources
▾
Log in
Sign up
01
02
03
04
05
01 / 05
·
Sears Catalogue · 1902
·
52 regions
High-precision document intelligence. On our infrastructure or yours.
GH
LI
X
HF
PRODUCT
Platform overview
Processors
Public Playground
Pricing
USE CASES
Model training data
RAG indexing
Agentic workflows
Audit & compliance
Web-ready PDFs
COMPANY
About
Careers
Customers
Contact
Blog
RESOURCES
Benchmarks
Open source
Security
Status
Documentation
© 2026 Datalab
Privacy
Terms of Use
End User License Agreement

[content]
High-Precision Document Parsing and Extraction API | Datalab
High-precision document intelligence.
We are a research lab that trains document intelligence models, trusted by frontier AI labs, museums, and enterprises where accuracy is non-negotiable.
Talk to sales →
Try playground for free →
Paste into your agent:
datalab.to/agent-setup.md
Copy
TRUSTED BY
1
HDR
2
H1
3
H1
4
TEXT
5
H1
6
TEXT
7
TEXT
8
H1
9
TEXT
10
TEXT
11
H1
12
TEXT
13
TEXT
14
H1
15
H1
16
IMG
17
TEXT
18
TEXT
19
H1
20
IMG
21
TEXT
22
TEXT
23
TEXT
24
TEXT
25
H1
26
TEXT
27
H1
28
TEXT
29
TEXT
30
H1
31
IMG
32
TEXT
33
TEXT
34
TEXT
35
TEXT
36
H1
37
TEXT
38
IMG
39
IMG
40
H1
41
TEXT
42
TEXT
43
TEXT
44
TEXT
45
H1
46
TEXT
47
H1
48
TEXT
49
H1
50
H1
51
TEXT
52
TEXT
CHANDRA · PARSE
Sears Catalogue · 1902 · 52 blocks
CONFIDENCE
96.9%
1
Page Header
Page Header
0.977
2
Section Header
Builders' Hardware
0.959
3
Section Header
THE GOODS WE LIST
0.984
4
Text
In this department are guaranteed for quality. They are made in proper proportions and are not only beautiful…
0.965
5
Section Header
THE FREIGHT RATES
0.980
+
47 more regions detected
chandra · v1.4.2 · 1,287 ms · markdown · json · html
[02]
PROCESSORS
Models deployed for any workflow
Each processor is rigorously benchmarked and designed for easy composition in your workflows.
P · 01
Convert
Use Chandra 2.5 to convert PDFs, spreadsheets, slides and more to markdown, HTML, and JSON, in fast, balanced, or accurate mode. Add word-level bounding boxes, chart data, and structured figure descriptions.
→
P · 02
Extract
Pull the fields you define into JSON, with citations and a confidence score on every field. Four modes, from turbo to accurate.
→
P · 03
Segment
Identify natural boundaries between documents that were originally combined into one file.
→
P · 04
Eval
Evaluate the quality of output, using our criteria or your own rubrics for quality.
→
Also on the platform: the document agent, form fill, form field detection, track changes, spreadsheet tables, and document creation.
See every processor and its price
.
[03]
BENCHMARKS · PUBLIC
Measured in the open.
OmniParseBench is our open benchmark: 16,288 pass/fail tests, every parser we could run, and every verdict public. Datalab accurate scores 94.1, ahead of Reducto, LlamaParse, Mistral OCR, Extend, and Azure.
Overall
OmniParseBench
94.1%
Datalab accurate
Datalab accurate
94.1%
Datalab balanced
92.6%
Reducto
90.8%
+3.3 vs the next vendor API
→
Tables
OmniParseBench · table tag
95.6%
Datalab accurate
Datalab accurate
95.6%
Extend
94.7%
LlamaParse
93.7%
#1 of 10 parsers
→
Multi-column
OmniParseBench · multi_column tag
94.9%
Datalab accurate
Datalab accurate
94.9%
Datalab balanced
92.7%
LlamaParse
91.4%
#1 of 10 parsers
→
olmOCR-bench
olmOCR-bench · Chandra 2.5
86.9%
Chandra 2.5
Chandra 2.5
86.9%
Chandra OCR 2
85.8%
LightOnOCR-3-4B
85.4%
+1.1 vs Chandra OCR 2
→
All benchmarks →
IN THEIR WORDS
←
"
We see Datalab as the bridge that helps us transform those old formats into something modern and easy to work with.
Gamma
Jon Noronha · Co-founder
CASE STUDY
→
"
It feels like magic. You do all the work, but we get the credit.
Purchaser.ai
Justin · Co-founder
CASE STUDY
→
"
Within 30 minutes we had a deployment ready to go. We haven't touched the implementation since.
RevisionDojo
Honza Kocourek · Founding Engineer · YC F24
CASE STUDY
→
"
We were quite blown away by the quality of extraction, especially on the type of complex academic papers that we were passing in.
Radical AI
Rohil Kulkarni
CASE STUDY
→
→
[04]
PLATFORM · CORE
Compose. Deploy. Monitor.
One pipeline runtime — from a single API call to 1B+ pages in a day.
01
COMPOSE
Wire processors into a pipeline.
Pick the processors you need. Compose them in the playground. Promote a versioned pipeline to production.
02
DEPLOY
Run our models anywhere.
Managed cloud, in your VPC, or fully air-gapped. We own the models — so you choose the network.
03
MONITOR
Monitor improvements and flag regressions.
Continuous evals against rubrics and a reference corpus — quality improves as we ship model updates, and any regression gets flagged.
[05]
VERTICALS · USE CASES
Trusted in countless mission-critical workflows.
Same engine — different documents. From training corpora to clinical trials to claims forms, every vertical gets the same bbox-precise parsing.
V · 01
AI Research Labs
V · 02
AI x Science
V · 03
Financial Services
V · 04
Healthcare
V · 05
Insurance
V · 06
Try your own
1
CAP
2
FIG
3
CAP
4
FIG
5
CAP
6
TEXT
7
TEXT
8
EQN
9
TEXT
10
TEXT
11
H1
12
TEXT
13
TEXT
14
FN
15
FTR
CHANDRA · PARSE
/
Attention Is All You Need · arXiv
15 blocks
/
96.9%
Training-grade corpora.
Parse web crawls, scientific papers, multilingual sources, and 1000+ page books into clean, citation-ready training data.
▸
Multilingual web corpora
▸
Scientific papers (PDF + LaTeX)
▸
Long-document books (1000+ pp)
▸
Math, charts, diagrams
"Working with Datalab has helped us move faster and solve complex edge cases with our documents. We need highly accurate document processing, and Datalab has been a key part of our data infrastructure. We're excited to continue to partner with Datalab."
— S., Member of Technical Staff, Frontier Lab
See AI Research Labs →
Talk to sales →
All customer stories →
[06]
MANAGED BATCH · NEW
1B+ pages in a day.
Zero infrastructure.
Burst at scale on our GPUs. We allocate the workers, you pay per page processed. In production at the largest AI labs and Fortune 500 enterprises.
Try Playground →
Talk to sales →
[07]
OPEN SOURCE · COMMUNITY
Open Source first.
Marker, Surya, Chandra, and Indra are our open source models — used by thousands of teams, 76k+ stars combined. Our managed platform handles scaling infrastructure, includes better proprietary models, and has enterprise-grade reliability.
Talk to sales →
Marker + Surya
High-throughput OCR engine for resource-constrained environments. Covers 90+ languages.
62k ★
Chandra
Our flagship open OCR model. Works on the messiest docs and the hardest edge cases.
12.6k ★
Indra
A 1M-parameter OCR model that reads handwriting, 80+ languages and pages at any angle, fast on a plain CPU.
[08]
DEPLOYMENT · MODES
Run it on our infrastructure. Or yours.
Because we own the models, you can run them anywhere — including networks that don't touch the internet.
01
Managed Cloud
Sign up, get an API key, ship in minutes. Free tier available.
Auto-scaling
SOC 2 Type II
99.99% uptime
Pay-as-you-go
Try Playground →
02
VPC Deployment
Datalab runs inside your private cloud. Compliance, residency, control.
AWS / GCP / Azure
Custom BAA / DPA
Sub-processor transparency
Dedicated support
Talk to sales →
03
On-prem · Air-gapped
For research labs and regulated environments where data cannot leave the network.
No internet required
Full model weights
Custom hardware support
White-glove deployment
Talk to sales →
[09]
BLOG · LATEST
What we shipped lately.
Model releases, benchmarks, and research notes from the team.
Product Updates
· Nov 6, 2026
API submits and result checks, 2x faster
Submitting a document now takes 314 ms at the median, down from 670 ms, and result checks are twice as fast. SDK 0.9.1 picks up finished results sooner.
Read →
Benchmarks
· Nov 4, 2026
Which document-parsing benchmarks can you trust?
We checked 15 document-parsing benchmarks: we ran seven open models through the 13 with public data, had three judge models count their real errors, and checked which benchmarks rank parsers the way their errors do.
Read →
Tutorials
· Nov 2, 2026
Data extraction vs parsing: Which to use?
Choose parsing for document context and extraction for specific fields. Extract, validate, and evaluate financial data with a JSON schema and the Datalab API.
Read →
All posts →
[10]
START
Start parsing.
Free tier available. No credit card required.
Try Playground →
Sign Up
Made by Datalab, an applied document intelligence R&D lab.
From Manhattan, NY.
❦
