iDBQuery vs Dataherald: what's the difference?
Dataherald is an open-source natural-language-to-SQL engine developers embed and configure themselves, whereas iDBQuery is a complete product that connects across databases, warehouses and files, cites every figure to its source row, generates charts and dashboards, and deploys in the cloud, your VPC or air-gapped — with no engine to host and tune.
Dataherald offers an open-source engine and API for converting natural-language questions into SQL, aimed at developers who want to add a text-to-SQL capability to their own application. You configure it, host it and build the surrounding experience. iDBQuery is that surrounding experience, delivered as a finished product.
- Product, not a component. Connect a source and ask — iDBQuery includes the interface, chart and dashboard generation, and shareable reports, so you're not building a front end around an engine.
- Cross-source answers. It builds one live model across MySQL, PostgreSQL, MongoDB, Excel, CSV, PDFs, REST APIs and SQL warehouses, letting a single question join multiple systems.
- Cited and verifiable. Every number traces to its exact source row, and you can inspect the SQL it wrote.
- Schema understanding built in. A semantic layer decodes cryptic schemas automatically, reducing the manual context-tuning that self-hosted engines usually require.
- Deploy anywhere. Cloud, VPC, fully air-gapped, Desktop app or Embedded SDK.
Dataherald is a solid fit if you're an engineering team building your own analytics feature and want an open NL-to-SQL core to embed. iDBQuery suits teams who want cited, cross-source answers now, without hosting an engine and constructing the product around it.
Updated 2026-06-22