Effective date: June 2, 2026
By accessing or using Sablier ("the Service"), including the sablier-flow Python SDK and the hosted compute service it calls, you agree to be bound by these Terms of Service. If you do not agree, do not use the Service.
Sablier provides synthetic alternative-history generation for financial-market data. Customers send a pandas.DataFrame of historical returns or prices via the sablier-flow Python SDK; we train a generative model on hosted GPUs and return synthetic paths that share the data's statistical fingerprint. The SDK is released under the Apache 2.0 license; the hosted compute backend is proprietary and accessed under these Terms.
You agree not to:
The Service is billed on a credit-based model. Credits are purchased in advance (via a subscription or a one-off pack) and consumed by sf.fit, sf.generate, and sf.validate calls at the rates published on sablier.ai. Subscriptions auto-renew at the end of each billing period unless cancelled in the dashboard. Credits are non-refundable except where required by law.
The Service generates synthetic financial-market data and statistical outputs (Sharpe ratios, overfit scores, predictive-rank scores, validation reports). None of this is investment, financial, legal, accounting, or tax advice. Synthetic paths are statistical samples, not forecasts. All trading and investment decisions are your sole responsibility. Past performance and synthetic counter-factual performance do not guarantee future results.
The Service is provided "as is" and "as available". To the maximum extent permitted by law, we disclaim all warranties, express or implied, including warranties of merchantability, fitness for a particular purpose, and non-infringement. We do not warrant that generative outputs will match any specific empirical property of real markets, that the Service will be uninterrupted or error-free, or that all defects will be corrected.
To the maximum extent permitted by law, Sablier will not be liable for any indirect, incidental, special, consequential, or punitive damages, or any loss of profits or trading losses, arising from or relating to your use of the Service. Our aggregate liability for any claim arising from these Terms will not exceed the amount you paid us in the twelve months preceding the event giving rise to the claim.
You retain all rights to the DataFrames you submit and to the synthetic outputs you receive. The sablier-flow SDK is licensed under Apache 2.0; you may use, modify, and redistribute it in accordance with that license. The hosted backend, the trained model weights, the FinBench benchmark, and the finval scoring library remain our property (the latter two are open-sourced under their own licenses).
You may terminate your account at any time from the dashboard. We may suspend or terminate your access for material breach of these Terms, with notice where reasonably possible. On termination your right to access the hosted backend ends; the Apache-licensed SDK you have installed remains yours under its license.
These Terms are governed by the laws of Italy, without regard to conflict-of-laws principles. Any dispute will be resolved in the competent courts of Milan, Italy, subject to any mandatory consumer protection rights you have under your local law.
We may update these Terms from time to time. Material changes will be announced via email or on this page with an updated effective date. Continued use of the Service after the effective date of a change constitutes acceptance of the revised Terms.
Questions about these Terms can be sent to team@sablier.it.