Skip to main content
Nori is a tabular foundation model for regression via in-context learning: give it a few labeled rows as context and it predicts new rows in a single forward pass — no task-specific training or fine-tuning. 
from synthefy_nori import NoriRegressor

model = NoriRegressor()          # downloads pretrained weights on first use
model.fit(X_train, y_train)      # "fit" just stores the labeled context — no training
preds = model.predict(X_test)    # predictions in a single forward pass

​
Choose your task

Tabular regression → Nori

Predict a numeric column from a table. No training, works on small data, uncertainty bands included.

Time-series forecasting

Multivariate forecasting with external context via the Synthefy API.

​
Why Nori

No training

Pretrained once; predict in seconds. No per-dataset training loop.

Works cold-start

Strong predictions from a handful of rows.

Uncertainty, free

A full predictive distribution — quantiles and prediction intervals out of the box.

Familiar API

A scikit-learn–style fit / predict estimator.

Explainable

SHAP, partial dependence, and feature selection built in.

Runs anywhere

Locally (open-source, GPU or CPU) or via the hosted API.

​
Get started

Install

Pick the right package in one table.

Quickstart

Fit and predict in five lines.

​
Need help?

API key

Set up a key for the hosted API.

Support

Email our team.