Search the web like a database.
Query and transform scattered web information into structured datasets. The open source alternative to Exa Websets.
From plain English to structured data
Dataembed turns a description of what you want into a verified, structured dataset. The open source alternative to Exa Websets.
Plan from plain English
Describe a webset in natural language. An agent turns it into verifiable criteria and enrichment columns before it searches.
Verified with cited evidence
Each candidate page is visited and checked against your criteria. Every claim links back to the evidence that supports it.
Live results, easy export
Rows stream into the data grid as they're verified. Export the finished webset as CSV or JSON.
Open source & self-hosted
Full control over your data and infrastructure. Deploy on your own servers or use the hosted version.
Queries that become websets
Every query is planned into verifiable criteria and enrichment columns before the search starts. Here's what that looks like.
"AI startups in San Francisco with Series A funding"
Criteria: headquartered in San Francisco, raised a Series A. Columns: company, funding amount, lead investor, website.
"Project management tools built for remote teams"
Criteria: project management product, designed for distributed teams. Columns: pricing, platforms, standout features.
"Recent research papers on machine learning in healthcare"
Criteria: peer-reviewed, applies ML to a clinical problem. Columns: title, authors, venue, publication date.
One query, every search provider
Each search fans out across multiple providers in parallel, and the ranked results merge with reciprocal rank fusion — so no single index limits what you find.
How the fan-out worksExa
Neural search tuned for finding entities, not just pages.
Parallel
A search API built for AI agents and deep research.
Firecrawl
Web search with clean, structured page content.
AI Gateway
Search-grounded models via the Vercel AI Gateway.
Choose your deployment option
Start with our free open source version or use our managed cloud service. Enterprise options available for custom deployments.
What's included
- Unlimited searches
- Full source code access
- Self-hosted deployment
- Community support
Managed Cloud
$29/month
Hosted for you, with higher limits and saved websets. Nothing to run yourself.
Get startedWhat's included
- Higher search limits
- Websets saved to your account
- CSV & JSON export
- Email support
What's included
- Unlimited searches
- Dedicated infrastructure
- SSO & audit logs
- Custom limits & terms
- Priority support & SLA