Downloads · 30 days
30
9% of all-time downloads
Qwen2.5-Omni-3B-cagatay-4bit
cagataydev/Qwen2.5-Omni-3B-cagatay-4bit
Qwen2.5-Omni-3B-cagatay-4bit is a text generation model from cagataydev. Use it when you need the model to write or continue text. It is set up for mlx. The card lists the license as apache-2.0.
A 4-bit quantized MLX model for Apple Silicon — fine-tuned for voice commands and robotics reasoning.
All-time downloads
330
Public
Parameters
3.1B
1.7 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors1.7 GB · 99%
Parameter types
How the weights are stored.
U323.1B · 100%
Try a prompt
Base models
Model card
From the Hugging Face model README
🌐 Q-Omni MLX — Qwen 2.5 3B Cagatay (4-bit)
A 4-bit quantized MLX model for Apple Silicon — fine-tuned for voice commands and robotics reasoning.
6.2 GB → 1.7 GB | 4-bit | Runs on MacBook Air
🚀 Use with Strands Agents + MLX
from strands import Agent
from strands_mlx import MLXModel
model = MLXModel(model_id="cagataydev/Qwen2.5-Omni-3B-cagatay-4bit")
agent = Agent(model=model)
agent("Listen to the voice command and plan the robot's next action")
📊 Model Details
| Property | Value |
|---|---|
| Base Model | Qwen/Qwen2.5-3B-Instruct |
| Fine-tune | cagataydev/qwen2.5-omni-3b-cagatay (LoRA) |
| Quantization | 4-bit MLX (group size 64) |
| Size | 1.7 GB |
| Platform | Apple Silicon (M1/M2/M3/M4) |
📦 Q-Model Family
| Model | Size | Quantized | Use Case |
|---|---|---|---|
| 🌐 Q-Omni (this) | 3B | 1.7 GB | Voice & multimodal |
| 🐤 Q-Tiny | 4B | 2.4 GB | Task planning |
| 🧠 Q-Brain | 35B MoE | — | Complex reasoning |
Built with DevDuck 🦆 and Strands Agents 🧬
- Task
- Text Generation
- Library
- mlx
- Type
- qwen2
- License
- apache-2.0
- Languages
- en
- Created
- Mar 28, 2026
- Updated
- Mar 28, 2026