Uncensored LLM providers with zero data retention
Four uncensored LLM providers in this directory offer zero data retention: Silk Compute, Abliteration, Featherless and Awan LLM. Morph keeps inputs up to 30 days on paid tiers, and ModelsLab's privacy policy doesn't mention prompts.
Updated Policies quoted word for word
| Provider | Policy | In their words | Note |
|---|---|---|---|
| Silk Compute | Zero data retention | "Silk Compute does not store your prompts or model outputs."Source | No moderation layer |
| Abliteration | Not stored by default | "Prompts, completions, and images are processed in memory and never stored by default."Source | Except web search and fetch |
| Featherless | Not logged | "Featherless does not log chats, prompts, or completions sent through our API."Source | |
| Awan LLM | Not logged | "We do not log user prompts and generations."Source | |
| Morph | Kept up to 30 days | "We keep Inputs and Outputs only to operate the Services and your own request history, delete them within 30 days, and do not review them in the ordinary course."Source | Flagged prompts kept on every tier |
| Unfil.AI | Temp data up to 24 h | "We do not store your API prompts or responses after processing"Source | Usage logs with IP addresses, 30 days |
| ModelsLab | Not stated | The privacy policy says nothing about prompts or outputs.Source |