API Terms
Version 2026-11-14
These terms cover the fross API only. You accept them by ticking the box before your first key — the acceptance, and this version, are recorded on your developer account.
1. What you may build
The API turns material you send into study content. You may use it for your own products, internal tools and client work.
Do not use it for anything illegal, harmful, deceptive or abusive: no content that incites violence or hatred, sexualises minors, gives unverified medical, legal or financial advice as if it were professional advice, or is designed to harass or defraud anyone.
2. Your keys
Keep keys secret and keep them on a server. A key in a browser or a mobile app is a key anyone can copy.
Do not resell raw keys, share a key between unrelated customers, or republish the API itself under your own brand as if it were your own model. Building a product on top of it is exactly what it is for.
You are responsible for everything done with your keys. Roll a key the moment you think it has leaked.
3. Limits and fair use
Each plan has hard server-side limits: generations a month, requests a minute, characters per set and items per generation. They are counted across all the keys on your account.
Do not try to get around them — multiple accounts to claim more free generations, automated key creation, or traffic designed to exhaust our capacity. We may throttle or suspend an account that does.
4. How your material is processed
The text you send is processed by third-party AI providers to build the study content, and by a second model to verify it. Do not send material you have no right to process, and do not send secrets, credentials or personal data you are not allowed to hand to a processor.
Sets created through the API belong to your key. They are not shown in the fross study app and are not readable by other developers.
5. Suspension and changes
We may suspend an account — and its keys stop working immediately — if it breaks these terms, abuses the limits, or creates risk for the platform or other users.
We may change these terms and the plan limits. A material change is announced by email before it applies. The version you accepted is recorded on your account.
6. No warranty
The API is provided as it is. AI output can be wrong; the source line and the second-model check exist to make it traceable, not to make it infallible. Check anything that matters before you rely on it.
To the extent the law allows, we are not liable for indirect or consequential loss arising from your use of the API.