The open engine, the shared knowledgebase. Run it on your own CPU.

Apache 2.0 code and open weights. Runs on CPU in your own cloud project.

Try it free
View on GitHub
Terminal cloning Prereasoner and deploying it to your own cloud project
Why us

The database does the math

The model proposes, the database computes

A 7B SQL model proposes queries. Each one is parsed and validated before it runs, and the database computes the answer.

The model proposes, the database computes
Deterministic and runs on CPU

Deterministic and runs on CPU

No GPU required. Model files are pinned by hash and decoding does not sample, so the same setup gives the same query.

Run it in your own cloud

Apache 2.0 engine and open weights. The models are pinned, so no vendor can change them under you. You pay your cloud provider for the infrastructure.

Run it in your own cloud
The shared knowledgebase

Typed world knowledge, built in the open

Grounded in real, shared sources

Geography and currencies come from public sources such as Wikidata and the ECB. They resolve a city to its country and currency. Schema.org supplies the vocabulary.

Grounded in real, shared sources
Help build the interpretability wiki

Help build the interpretability wiki

schema.org defines the classes of the world. The data that fills them is scattered and mostly closed. We want to build a community wiki of open reference datasets. Contributors welcome.

Where the line sits

Community Edition resolves the world. Structured RAG resolves your world

Community EditionStructured RAG
The engineOpen sourceIncluded
Open weights on Hugging FaceYesYes
Run on your own CPUYesYes
Global knowledgebase (geo, forex, schema.org)Downloaded by the deploy scriptHosted
Who runs itYou, in your cloudWe do
SetupGuided deploy scriptNone

Do not want to run it yourself? Use Structured RAG.

Deploy it yourself

Clone, fetch, deploy

01 CLONE

Get it

Clone the Apache 2.0 repository. python -m engine.fetch_weights pulls the open weights. No account or token needed.

02 DEPLOY

One script

The deploy script sets up the services and a Postgres database in your Google Cloud project. Billing must be enabled.

03 CONNECT

Your data

Connect spreadsheets or CSV files.

04 QUERY

Auditable answers

Return the result with its query and source rows.

Open source

The open engine, the shared knowledgebase. Run it on your own CPU

Apache 2.0 code and open weights, in your own cloud.

Try it free
Source on GitHub Weights on Hugging Face