SillyTavern

SillyTavern

Shares roleplay presets for SillyTavern, mainly the Voyage series plus a generic Eval preset. It suits people who run story chats with local or hosted language models.

Made by nohurry on Hugging Face. These notes are Mindket's, written from its model page and model card; the model's own files and license are on Hugging Face.

About

What it is: This is a public collection of presets for SillyTavern, a chat front end for roleplay with language models. Its main series is Voyage, shown in versions from v2.0 to v6. The intro also names lorebooks and characters, though the visible text covers presets.

Who it suits: It suits roleplay fans who already use SillyTavern and want ready-made prompt setups. Preset writers can also read the files to see how others structure rules.

Industry | Who uses it | How they use it | Example
Hobby and games | Roleplay fans | Load a Voyage preset to run story scenes with NPCs | Using v6 for its scene tracker and news system
Local AI | Users of Gemma 4 or Mistral models | Pair a preset with one of the models the author lists | Running v5.1 with a Mistral Small 3.2 model
Prompt writing | Preset authors | Read the Eval preset to try out rules | Experimenting with the rule playground

People who want a general chat assistant instead of roleplay can skip it.

Where to use it:

• SillyTavern, where you load the preset files
• Your own computer on Windows, macOS or Linux, where you download the files
• Models the author lists, such as Mistral Small 3.2, Gemma 4, DeepSeek and GLM

Use case | Who | What they get back
Story scenes with NPCs | Roleplay fans | A preset that guides NPC behavior and world events
Tracking a scene | Long-chat users | A scene tracker added by Voyage v6
Testing prompt rules | Preset authors | A playground of rules in the Eval preset

How to set it up:

• Python, for the download command
• A working SillyTavern install, which the card does not explain
• A model to chat with. The author tested with a 31B Gemma 4 build in a 4-bit quantization

1. Install the Hugging Face hub package.
2. Download the files in this repository.
3. Open SillyTavern and load the preset file you want from the Presets folder.

pip install -U huggingface_hub

hf download nohurry/sillytavern

How to use it: Pick a preset version that matches your model, load it in SillyTavern, and start a chat. The newest Voyage, v6, is described as working with any model released after 2024. Older versions list specific models.

Tool | What it does
Voyage v6 | Adds a news system, a scene tracker and an online-offline simulation, and drops PbtA rolling for now
Voyage v5.1 | Improves chain of thought, NPC interactions and user agency, and cuts looping and grammar mistakes
Voyage v5.0 | Reduces token use and improves world dynamics and NPC interactions
Eval v1.0 | A playground of rules for experiments, meant more for development than normal use

Example: You load Voyage v6 and start a long roleplay with a Gemma 4 model. The card says the preset then adds a scene tracker and a news system to your story. It does not show a sample chat.

For newcomers and experts: A preset is a saved set of instructions that SillyTavern sends to the model along with your messages. A first small step is to download the files, then load the newest Voyage preset and chat with a model from the author's list.

The series moved from cause-and-effect generation in v2.0 to a rewrite in positive prompts in v5.0. The author tested the presets with a Gemma 4 31B QAT build, MTP enabled and a 32K BF16 context. Results on other models may vary, since the card says only that other models might work.

Pick a preset version
Load its file in SillyTavern
Choose a model from the list
Chat with the preset's rules active

Good to know:

• License: MIT.
• Cost to run: none named. You supply your own model, and the card names no key or paid service.
• Data it reaches: it downloads files from the Hugging Face Hub. The card names no account or upload.
• Not for: the Eval preset, which the author calls a development resource more than a normal-use preset.
• Version 4.0 is labeled experimental.

On Hugging Face, 13 November 2026

• Downloads in the last 30 days: 0
• Likes: 58
• License: mit
• Languages: en

Links

• Model: https://huggingface.co/nohurry/sillytavern
• Files: https://huggingface.co/nohurry/sillytavern/tree/main

Get the model

Download its files with Hugging Face's command line:

pip install -U huggingface_hub
hf download nohurry/sillytavern

Then follow the usage in its model card. Nothing in this kit runs by itself.