FAQ
Quick answers to the questions that come up the most. If you need a step-by-step checklist, start with Troubleshooting.
General
What is IntenseRP Next v2?
A local desktop app + OpenAI-compatible API that lets clients like SillyTavern talk to provider web chats (DeepSeek / GLM / Moonshot / QwenLM / Perplexity / HuggingChat / Google AI Studio) by driving them in a real browser.
Is this an official DeepSeek or SillyTavern project?
Nope. This is a third-party tool and is not affiliated with DeepSeek, SillyTavern, or any provider.
Do I need a paid provider API key?
No. IntenseRP uses provider web apps you log into in the browser (DeepSeek / GLM / Moonshot / QwenLM / Perplexity / HuggingChat / Google AI Studio). This project does not use paid official provider APIs.
What platforms are supported?
Windows and Linux are the supported targets right now. A graphical desktop is required because it launches a real browser.
Why does it open a browser window?
Because that's how it works: IntenseRP actually drives the provider UI (and intercepts the underlying network stream). It is not a headless, server-only API.
Setup & usage
What endpoint do I paste into SillyTavern?
By default:
http://127.0.0.1:7777/v1
If you change the port in Settings, update the endpoint to match.
Do I need to keep the browser open?
Yes - keep the provider browser window open while you use IntenseRP. You can minimize it, but if you close it you'll usually need to press Start again.
Which model should I pick?
Use the preset that matches your active provider:
- DeepSeek:
deepseek-auto/deepseek-chat/deepseek-reasoner - GLM Chat:
glm-auto/glm-chat/glm-reasoner - Moonshot:
moonshot-auto/moonshot-chat/moonshot-reasoner - QwenLM:
qwen-auto/qwen-chat/qwen-reasoner - Perplexity:
perplexity-auto/perplexity-chat/perplexity-reasoner - HuggingChat:
huggingchat-auto/huggingchat-chat/huggingchat-reasoner - Google AI Studio:
aistudio-auto/aistudio-chat/aistudio-reasoner
Why are responses slow sometimes?
Two common reasons:
- IntenseRP processes one request at a time by default (requests are queued)
- DeepThink can add extra time
For details, see API Behavior.
My character names look wrong (or always say User/Character).
This is usually just name passing/detection. Start here: Formatting.
Networking & security
Can I connect from my phone or another PC?
Yes - enable Allow Local Network Access, then use your computer's LAN IP in the endpoint (example: http://192.168.1.100:7777/v1). See Network & API.
Can I expose IntenseRP to the internet?
Not recommended. IntenseRP is designed for local/LAN use. If you need remote access, use a VPN and consider enabling API keys.
Why am I getting 401 Unauthorized?
You most likely enabled Require API Keys in Settings. Either disable it, or put the same key into your client. See API Keys.
Privacy & data
Does IntenseRP store my provider password?
Only if you enable Sign In Automatically. Your config directory contains sensitive data (credentials, API keys, cookies), so treat it like a password vault. If you prefer, leave it off and just log in manually.
Where are my settings and sessions saved?
In your config directory ([config_dir]). If you enabled Persistent Sessions, it can also contain saved browser profiles for your providers. See System.
Troubleshooting & support
The first launch takes forever - is it frozen?
Usually not. On first run it may download/verify browser components. Let it finish once - subsequent launches are much faster.
What's the best way to report a bug?
GitHub issues are best. Include your version, OS, exact steps to reproduce, and logs (redacted). Start with Troubleshooting for a checklist and a copy/paste template.
I still need help - where do I reach you?
For quick help and shared troubleshooting, join the Discord server.
For private contact options, see Means of Contact.