# SillyTavern-Character-Creator

**Repository Path**: ffpy001/SillyTavern-Character-Creator

## Basic Information

- **Project Name**: SillyTavern-Character-Creator
- **Description**: No description available
- **Primary Language**: Unknown
- **License**: MIT
- **Default Branch**: main
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2025-09-25
- **Last Updated**: 2025-09-25

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# SillyTavern Character Creator (CREC)

## Overview

A [SillyTavern](https://docs.sillytavern.app/) extension that helps you create character cards based on your ST data with LLMs using [connection profiles](https://docs.sillytavern.app/usage/core-concepts/connection-profiles/).

![popup](images/popup.png)

---

![settings](images/settings.png)

---

**If you are using a _Text Completion_ profile, make sure your profile contains API, preset, model, and instruct.**

**If you are using a _Chat Completion_ profile; API, settings, model would be enough.**

---

## Installation

Install via the SillyTavern extension installer:

```txt
https://github.com/bmen25124/SillyTavern-Character-Creator
```

To open the CREC popup, click the extension icon:

![icon](images/icon.png)

## Demo Video

https://github.com/user-attachments/assets/4ed6fbb3-c2a4-4cdc-8692-406af9094266

## FAQ

>Can I use this with my local 8B/12B RP model?

Most likely, yes. If you can't, try changing _Output Format_.

>Can you suggest a model?

Gemini models are cheap, fast, and efficient. I usually use Gemini Flash 2.0. But most models should work fine.

>What is the difference compared to alternatives?

In general, alternatives are just websites. This means you can't feed the AI with your ST character/lorebook data. They mostly use a single model. Their customization is limited.

>What is the difference compared to [chargen](https://chargen.kubes-lab.com/)?

The one thing chargen might be better is it can give better results because it uses [chargen-v2](https://huggingface.co/kubernetes-bad/chargen-v2) model that trained from character cards.  But since CROC is customizable, you can even use _chargen-v2_ on your local.

>What is the difference compared to [pookies](https://pookies.ai/create)?

There are 2 advantages of pookie. 1. You can give a fandom website so it can analyze it. 2. It has detailed fields like _age, gender, running outfit_. Currently, I'm not planning to implement detailed fields because their quality differs from LLM to LLM.