Sign in Subscribe

How to Install SillyTavern and Connect Your First AI Model (Mac/Windows/Linux complete guide)

Tracy Giang

• 5 min read
Share
How to Install SillyTavern and Connect Your First AI Model (Mac/Windows/Linux complete guide)
How to Install SillyTavern and Connect Your First AI Model (Mac/Windows/Linux complete guide)


What is SillyTavern?

SillyTavern is a free, open-source local frontend for chatting with AI models through character cards, custom prompts, lorebooks, and more. It is not an AI itself — you install it, then connect it to a cloud API or a local model backend. SillyTavern offers two branches: release (recommended for most users, updated roughly monthly) and staging (latest features but may break at any time, for power users only).

Prerequisites (All Platforms)

You need two basic tools: Node.js (version 20 or newer) and Git. Node.js 20+ is required because current SillyTavern builds expect a modern Node environment. Git is used to download and update the official project files from GitHub.

You can verify they're installed by opening a terminal and typing:

node --version

git --version

If both return version numbers, you're ready. If not, install the missing tool first.

🪟 Windows Installation

Important warnings:

  • Do NOT install into any Windows-controlled folder (Program Files, System32, etc.)
  • Do NOT run Start.bat with admin permissions
  • Windows 7 cannot run Node.js 20 and is unsupported
  1. Install Node.js (latest LTS version is recommended)
  2. Install Git for Windows
  3. Open Windows Explorer (Win+E) and browse to or create a folder you control (e.g. C:\MyAIFolder)
  4. Click the address bar at the top, type cmd, and press Enter
  5. In the Command Prompt, run: git clone https://github.com/SillyTavern/SillyTavern -b release
  6. Once cloned, double-click Start.bat — Node.js will install all requirements automatically
  7. The server starts and SillyTavern opens in your browser

Method 2: Via SillyTavern Launcher (Easiest)

Press Windows + R and run:

cmd /c winget install -e --id Git.Git

Then press Windows + E, navigate to your desired folder, type cmd in the address bar, and run:

git clone https://github.com/SillyTavern/SillyTavern-Launcher.git && cd SillyTavern-Launcher && start installer.bat

🍎 macOS Installation

All steps are done in the Terminal app.

Using the SillyTavern Launcher (recommended for Mac):

  1. Install Homebrew: /bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh)"
  2. Install Git: brew install git
  3. Download the launcher: git clone https://github.com/SillyTavern/SillyTavern-Launcher.git
  4. Navigate into it: cd SillyTavern-Launcher
  5. Run the install script: chmod +x install.sh && ./install.sh — choose what you want to install
  6. After installation, start it with: chmod +x launcher.sh && ./launcher.sh

Manual Git install (alternative):

git clone https://github.com/SillyTavern/SillyTavern -b release

cd SillyTavern

bash start.sh

🐧 Linux Installation

Manual Git install:

  1. Install Git and Node.js 20+ via your distribution's package manager
  2. Clone the repo: git clone https://github.com/SillyTavern/SillyTavern -b release
  3. Navigate into it: cd SillyTavern
  4. Run the start script: ./start.sh

If your distro ships an outdated Node.js package, use an official Node version manager (like nvm) or the NodeSource repository to ensure you're on Node.js 20 or newer.

Via Launcher:

git clone https://github.com/SillyTavern/SillyTavern-Launcher.git

cd SillyTavern-Launcher

chmod +x install.sh && ./install.sh

chmod +x launcher.sh && ./launcher.sh

⚠️ Important: Keep the Terminal Open

Do not close the terminal window once the browser opens. That window is running the local server — if it closes, the page stops responding.
SillyTavern runs at http://localhost:8000 by default.

Connecting Your First AI Model

Once SillyTavern is open in your browser, click the API Connections button (the plug icon in the top bar).

SillyTavern can connect to a wide range of LLM APIs. Here are the most popular options:

☁️ Cloud APIs (Easiest for Beginners)

These require no heavy hardware — any modern PC will do.

Provider

Notes

MegaNova AI

A multi-model inference platform that aggregates 100+ models into a single, unified API. You pay per token to run cutting-edge models.

OpenAI

Easy to set up; requires prepayment for credits; newer models support multimodality

Claude (Anthropic)

Recommended for creative, unique writing style; requires prepayment for credits

Google AI Studio

Has a free tier with rate limits (Gemini Flash)

AI Horde

Works out of the box with no additional settings; uses volunteer GPUs; free but generation times vary

Pollinations

Requires no setup and can be used out of the box; access to a wide range of models free of charge

DeepSeek

Access to DeepSeek V3 and R1 models; requires a minimum $2 credit purchase but very cheap for the quality

💻 Local APIs (Privacy-Focused)

These run entirely on your own machine. They are free to use and have no content filter, but require separate download of LLM model files from HuggingFace (which can be 5–50 GB each).

Backend

Notes

Ollama

Easiest to set up; has a catalog of models available for one-click download

KoboldCpp

Easy-to-use with CPU offloading (helpful for low VRAM users); runs from a single binary; supports GGUF models

LM Studio

Popular GUI app — download it separately, load a model, and point SillyTavern to its local endpoint

Quick Connection Steps (Cloud Example: Meganova)

Meganova is a multi-model inference platform that allows you to pay per token to run over 100+ top AI models on a single platform with one unified API.

  1. Go to meganova.ai create an account generate an API key.
  2. In SillyTavern, click the API Connections (plug) icon.
  3. Set API Type to Chat Completion.
  4. Set Chat Completion Source to Meganova
  5. Paste your Meganova API Key.
  6. Select your desired model from the dropdown.
    • Text/Roleplay: GPT-5.5, Claude Opus 4.7, DeepSeek V4, Gemini, Qwen, Kimi...
    • Media/Generation: Wan video, Veo, image generation...
  7. Click Connect — you should see a green "Connected" indicator.
  8. Start a new chat and send a test message!
Click here for the detailed guide: "How to Set Up a Free & Private API for SillyTavern Using MegaNova AI (2026)".

Common Troubleshooting

Here are the most common issues and fixes:

  • "Node is not recognized": Restart the terminal after installing Node.js; reinstall if the version command still fails
  • Browser page won't open: Check whether the local server is still running; copy the local address from the terminal manually
  • Interface opens but no AI replies: Recheck the provider, endpoint, API key, and selected model
  • Windows blocks files: Move the install folder out of protected system locations and use a normal user-controlled directory
  • Local models are too slow: Try a smaller model or use a cloud API until you upgrade hardware

Tips for First-Timers

  • Use the release branch — it's the stable one and updates roughly once a month
  • Start with a cloud API like MegaNova AI before setting up local models
  • Once you have a working connection, import one character card and run a short conversation before adding extensions, custom themes, or complex prompt chains

The full official docs live at docs.sillytavern.app — it's well-maintained and the best reference for advanced features.

What’s Next?

Ready to upgrade your AI roleplay experience?

  • Sign up and explore now.
  • 🔍 Learn more: Visit our blog and documents for more insights or schedule a demo to optimize your search solutions.
  • Join the MegaNova community for the latest endpoint updates and technical support

Stay Connected

Sign up for more like this.

Enter your email
Subscribe