Snapshot
Continue to About ↓- What it does
- Nori is an open-weight foundation model for tabular prediction: you pass labeled example rows alongside the rows you want predicted, and it returns predictions without training or tuning a task-specific model.
More detail
- Synthefy offers it as free Apache 2.0 code and weights you run yourself, or as a hosted API and no-code platform.
- Best for
- Enterprise data science teams with many recurring prediction tasks (demand, revenue, churn, customer value) who want to skip per-task training and can run Python or use a hosted API.
More detail
- Strong fit for privacy-sensitive teams, since the open weights run locally on your own CPU or GPU.
- Pricing
- Two paths with different economics.
More detail
- The open-source model (synthefy-nori code and weights) is Apache 2.0 and free for commercial use, but you pay your own compute and infrastructure for local inference.
- Pricing for the hosted API, the Synthefy Platform, and Snowflake/SageMaker deployments is not pub
- Main limitation
- Teams needing text, image or video generation, autonomous agents, or classification-style GenAI — Nori is a tabular regression and time-series forecasting model.
More detail
- Also a poor fit for buyers who need published self-serve pricing or HIPAA coverage without an eligible deployment and signed BAA.
- Record
- Living and monitored
- Reviewed
- Evidence
- Public web research
- Machine access
- Structured profile ↗
About
- Nori is an open-weight foundation model for tabular prediction: you pass labeled example rows alongside the rows you want predicted, and it returns predictions without training or tuning a task-specific model.
- Synthefy offers it as free Apache 2.0 code and weights you run yourself, or as a hosted API and no-code platform.
What it helps with
- Data teams spend weeks engineering features, training and tuning a separate model for every prediction task
- Little reuse of a single model across a team's many prediction tasks
- Models miss relevant external data that could improve predictions
- Business stakeholders cannot see which inputs drive a prediction up or down
- Teams without GPU infrastructure or MLOps capacity to host and serve models
- Privacy-sensitive data that cannot leave a team's own environment
Capabilities
A practical view of what is available and where limitations apply.
| Offering | Available | Detail |
|---|---|---|
| No task-specific training | Yes | predict() takes test rows plus labeled context in one call; no fitting, hyperparameter search or fine-tuning per task |
| Open-source code and weights (Apache 2.0) | Yes | Free for commercial use; available on GitHub and Hugging Face; weights cached locally after first download |
| Local execution on private data | Yes | Runs via pip install synthefy-nori on CPU or single GPU with automatic CPU fallback |
| Managed hosted API | Yes | Hosted on Baseten with SOC 2 Type II, HIPAA-eligible and GDPR support; no model hosting or GPU infrastructure to manage |
| Interpretability | Yes | Built-in SHAP, partial dependence and feature selection, plus a Nori-Causal write-up on distinguishing drivers from followers |
| Uncertainty quantification | Yes | Full predictive distribution with quantiles and prediction intervals out of the box, per docs |
| Tables plus text input | Yes | Product page lists structured and unstructured data together |
| Time-series forecasting | Yes | Separate NoriTSForecaster package documented; setup guide available |
| Published benchmark wins over tuned XGBoost/LightGBM | Limited | Vendor-run evaluation on TabArena, OpenML-CTR23 and TALENT; no independent benchmark or review found |
| Nori Flash | No | Listed on the site as coming soon |
Bot trust
How confidently AI systems can verify this profile from public evidence.
—of 100
ChatGPT—/100
Claude—/100
This profile has not completed a Bot Score assessment yet.
Evidence
Where this record came from, when it was reviewed, and how confidently it can be used.
- Verification
- Scribble researched these facts from the linked public sources.
- The owner has not manually confirmed every field.
- Last reviewed
- Primary source
- synthefy.com
Technical profile details
Record ID
e56be9c3-70c7-4de9-af30-3a6f2212a6deCanonical URLhttps://profiles.scribble.network/listing/synthefy-e56be9c3