Quick Start
This guide helps you get from a fresh Tavern Studio install to your first character chat. The basic flow is simple: open the app, connect a model backend, import a character card, and start chatting.
Who this is for
- New users opening Tavern Studio for the first time.
- Users who are unsure how to configure an API key or local model.
- Users who already have a PNG character card and want to chat with it.
What you will learn
- How to find the main setup areas.
- How to configure your first cloud API or local LLM route.
- How to import a character card.
- How to start your first AI character chat.
Step 1: Open Tavern Studio
When you first launch Tavern Studio, you start with a clean local workspace.
- Language: if the interface is not in your preferred language, open Settings and change the language option.
- Character library: your character list may be empty at first. After importing cards, characters appear in the library.
Step 2: Connect a model backend
To make a character respond, Tavern Studio needs a model backend. The quickest option is usually a cloud API.
- Open Settings and find the API section.
- Choose your provider. For many users, OpenAI-compatible API is the most flexible option because many services support that format.
- Enter your Base URL.
- Enter your API Key.
- Refresh the model list.
- Choose the model you want to use and save or connect.
If you want to run fully local models, see Local GGUF Model Guide.
Step 3: Import your first character card
Tavern Studio supports common Tavern-style character cards.
- Prepare a PNG character card that contains character metadata.
- Import it by dragging the file into the app, or use the + import entry.
- After import, the character appears in your character library.
Step 4: Start your first chat
Open the imported character from the library. Tavern Studio will open a chat window for that character.
- Review the greeting message if the card includes one.
- Type your first message in the input area.
- Send the message and wait for the model response.
- If the response stops early, use continue generation when available.
Explore next
After your first chat, you can improve the experience by binding a world book, changing presets, or trying another model.
FAQ
Why does sending a message fail?
Check your API key, Base URL, model name, network connection, and account quota.
Why does my character card import as blank?
The image may not contain character metadata, or the metadata may have been stripped by social media compression. Try importing the original file.
Why was the reply cut off?
The model may have reached the response token limit. Use continue generation or increase the response length in your preset.
Next steps
- Interface Overview: learn where the main actions are.
- Configure Your First API: set up model connections in detail.
- Import Character Cards: learn more import and conflict handling.