An uncensored APIfor SillyTavern

SillyTavern connects to any OpenAI-compatible endpoint as a custom Chat Completion source. With a model that does not refuse, the jailbreak prompt can come out of your preset, and the model follows the character card instead of arguing with it.

Set it up

Create a key in the dashboard.
In SillyTavern open API Connections and fill in the fields as shown.
Press Connect, then pick a model. Remove any jailbreak text from your preset.

API Connections

API:                    Chat Completion
Chat Completion Source: Custom (OpenAI-compatible)
Custom Endpoint:        https://uncensoredopencode.com/v1
Custom API Key:         sk-nc-...
Model ID:               glm-5.3-flash-uncensored

Good to know

A million tokens of context changes how a long roleplay behaves: the model can keep the whole chat and a large lorebook in view rather than a summary of them. Set the context size in SillyTavern to match the model you picked.

For story work, Kimi K3.1 Pro writes the best prose on the line and Qwen3.8 27B is the cheap one for long casual chats.

Any id on the models page works here, and each is priced by the token.

Other harnesses

opencodeThe provider block for opencode.json.Hermes AgentUncensored models for Hermes Agent Claude CodeClaude Code on uncensored models CodexUncensored models in Codex

Bring your agent.

Get an API key Set up opencode See the models