DashboardAPI
Documentation

SillyTavern

Use NavyAI as the backend for SillyTavern. Point SillyTavern's Chat Completion API at NavyAI's OpenAI-compatible endpoint with your NavyAI key, then pick any model from the catalog.

Connect via Chat Completion (recommended)

  1. Open SillyTavern and click the API Connections (plug) icon
  2. Set API to Chat Completion
  3. Set Chat Completion Source to Custom (OpenAI-compatible)
  4. Set Custom Endpoint (Base URL) to https://api.navy/v1
  5. Paste your NavyAI key (sk-navy-...) into the Custom API Key field
  6. Click Connect — available models load automatically; pick one from the Model dropdown

Anthropic-native alternative

If you prefer SillyTavern's Claude chat completion source, enable its Reverse Proxy option, set the proxy URL to https://api.navy and use your NavyAI key as the proxy password. Both routes hit the same models — the Custom (OpenAI-compatible) source is the simpler default.

Picking a model

Any model in the NavyAI catalog works. Popular roleplay picks include claude-sonnet-4.6 and claude-opus-4.6 for prose quality, gemini-2.5-pro for long context, and claude-haiku-4.5 for fast, cheap swipes. Browse everything with:

Bash
1curl https://api.navy/v1/models

Premium models are marked with premium and required_plan in the catalog — the dashboard shows which tier your subscription unlocks.

Tips

  • Enable Streaming in SillyTavern for token-by-token replies
  • Set SillyTavern's context size to match the model's context_window from the catalog so long chats aren't truncated early
  • Existing prompts and presets work unchanged — NavyAI is a transparent OpenAI-compatible backend
  • If you hit 429, you've reached your plan's per-minute or daily limit; check GET /v1/usage or the dashboard. See the Errors & Rate Limits page for details.
← PreviousAPI EndpointsNext →Janitor AI