SillyTavern integration
Connect SillyTavern to Modellane with the custom OpenAI-compatible chat completion source and choose a model for your characters.
SillyTavern can send chats to any OpenAI-compatible endpoint. Use the custom chat completion source with the Modellane base URL and your key.
SillyTavern is a chat front end for characters and roleplay. It can send chats to any OpenAI-compatible endpoint through its custom chat completion source, which calls the Chat Completions API (POST /v1/chat/completions).
Connect#
- Open API Connections (the plug icon in the top bar).
- Set API to Chat Completion and Chat Completion Source to Custom (OpenAI-compatible).
- Fill in the fields:
| Param | Value |
|---|---|
Custom Endpoint (Base URL) | https://usemodellane.com/v1, without /chat/completions at the end |
Custom API Key | A key from API keys |
Model ID | lane-1, or any id from Models & pricing |
- Click Connect. When the status turns green, the model list is loaded and you can pick a model from it.
Verify#
Click Test Message in the connection panel, or send a short message to any character. A reply means the connection works. An authentication error points at the key, and model_not_found at the model id.
Recommended settings#
- Context size: set the context slider in the generation settings to the model's context length or lower. Long chat histories are sent with every message, so a smaller context also keeps each request cheaper.
- Max response length: this becomes
max_tokens. A lower value keeps the balance hold for each message small; see Deduction rules. - Streaming: turn it on to see replies as they are written.
- Prompt Post-Processing: if a model handles your chat layout poorly, for example several system messages in a row, try one of the post-processing options in the connection panel.
For writing personas and keeping long sessions consistent, see System prompts & roleplay. The model we recommend for roleplay, deepseek-roleplay, is a good place to start.