Effective 22 October 2026
These terms govern use of the inference API at api.tascoinference.com (the "Service"), operated by Fulvio Luca Daniele Ventura, VAT IT01772860852, Via Lorenzo Perosi 25, 93100 Caltanissetta, Italy ("we"). By sending requests to the Service you accept these terms. If you reach the Service through a platform such as OpenRouter, your relationship with that platform is governed by its own terms, and these terms apply between us and the platform.
The Service runs open-weight language models and returns their outputs through an OpenAI-compatible API. The models, prices, context limits and supported parameters on offer are those listed at api.tascoinference.com/v1/models at the time of each request. We serve the official weights released by the model authors, without modification. We may add or withdraw models and change prices; changes apply to requests sent after they are published.
You keep all rights to the prompts you send and to the outputs you receive. You grant us only the limited right to process them in order to answer your request. We do not store your content and we do not use it to train or improve models, as described in our Privacy Policy. You are responsible for having the rights needed to send your content to us.
You must not use the Service:
We may suspend access that breaches these rules.
Outputs are generated automatically by machine learning models. They may be inaccurate, incomplete or inappropriate, and they do not represent our views. You are responsible for evaluating outputs before relying on them, in particular for medical, legal, financial or safety-critical decisions.
Each model is distributed by its authors under its own license (for example, Qwen3.6 35B A3B is released under the Apache License 2.0). Your use of the outputs remains subject to any terms of that license that apply to them.
We aim for high availability but do not guarantee uninterrupted service. When capacity is full or a model is temporarily unavailable, the Service answers with HTTP 429 so that requests can be routed elsewhere. Unless a separate written agreement says otherwise, the Service is provided without a service-level commitment.
Requests are charged per token at the prices listed for the model, based on the token counts reported in each response. When you use the Service through a platform, you pay the platform under its terms, and the platform pays us.
To the extent permitted by law, the Service is provided "as is" and "as available", without warranties of any kind, express or implied, including fitness for a particular purpose and non-infringement.
To the extent permitted by law, we are not liable for indirect or consequential damages, lost profits or lost data, and our total liability for any claim relating to the Service is limited to the fees paid to us for the Service in the three months before the event giving rise to the claim. Nothing in these terms limits liability for willful misconduct or gross negligence, or any liability that cannot be limited under applicable law, including the rights of consumers.
We may update these terms by publishing a new version on this page with a new effective date. The new version applies to requests sent after that date.
These terms are governed by Italian law. Courts of Caltanissetta, Italy, have jurisdiction, without prejudice to mandatory rules protecting consumers.