Skip to main content
The brain of your toy is any LLM from the mlx-community organization on Hugging Face. OpenToys loads it with mlx-lm, and falls back to mlx-vlm for models that need it.

​
Models

Ministral 3

Default. Small and fast.

Gemma

Google’s open models.

Qwen3

Multilingual from Alibaba.

Llama

Meta’s open models.

gpt-oss

OpenAI’s open-weight model.

Muse Glimmer

Large 30B model.

​
Switching models

Open the Models page in OpenToys, pick an mlx-community repo and the app hot-swaps it without restarting your session.
Smaller 4-bit models respond faster, which matters for a toy. Start with a 3B model and size up if your Mac has memory to spare.

​
Personalities

Each character is a system prompt plus a voice. The prompt is what makes a toy a storyteller, a game host or a tutor. Default characters live in personalities.json.