Flovo AI
Privacy
Local by design
Roadmap prompts are processed by the selected local model in your browser. Flovo does not operate an AI inference server and does not require an API key.
When you explicitly choose a model, runtime and model files are fetched from their distribution hosts, including Hugging Face and jsDelivr. Those hosts receive normal network-request information required to deliver those files. Prompt text is not sent to those hosts as an inference request.
Qwen model assets are cached through Transformers.js. Gemma 4 LiteRT-LM files are stored in Flovo's browser CacheStorage when possible. Flovo requests persistent browser storage on a best-effort basis; your browser still controls quota and eviction.
Roadmap history is stored in browser storage. Flovo does not upload that history to an application server.
Open Local AI to see downloaded model state and remove a selected model. Clearing site data can also remove local history, application caches, and model files.