Run SillyTavern on Mingles Router
SillyTavern's “Custom (OpenAI-compatible)” chat completion source takes an endpoint URL, a key and a model id. Long conversations resend the whole history every turn — input price is everything here.
Pick your model and paste your key — every command and value on this page updates live.
Endpoint: https://router.mingles.ai/v1 · key stays in your browser ·
no key entered? commands show <your-key>.
Connect
1. Open API Connections
The plug icon → API: Chat Completion → Chat Completion Source: Custom (OpenAI-compatible).
2. Endpoint, key, model
Custom Endpoint = the base URL below; paste the key; set Model ID to {{model}}. Enable streaming.
3. Mind the history
Every turn resends the conversation, so input tokens dominate the bill — MiniMax at $0.18/1M input is built for exactly this shape of traffic.
Exact values
| API | Chat Completion |
| Chat Completion Source | Custom (OpenAI-compatible) |
| Custom Endpoint (Base URL) | https://router.mingles.ai/v1 |
| API Key / Model ID | <your-key> / deepseek-ai/DeepSeek-V4-Flash-0731 |
Switch model
Swap Model ID per character/preset — same connection.
Before you file a bug: read the limits
Reasoning models spend the output budget on internal thinking, output is capped at 8192 tokens, there is no KV cache, and there are no built-in web tools. Most “it broke on Mingles Router” reports are one of these. See Model limits & behavior →
Frequently asked
Card-free payment? +
Yes — USDT/USDC top-ups, no KYC. The free tier is enough to test a long conversation before paying anything.