Downloads · 30 days
0
Qwen3-30B-A3B-Instruct-2507
basecompute/Qwen3-30B-A3B-Instruct-2507
Qwen3-30B-A3B-Instruct-2507 is a machine learning model from basecompute. Use it for the machine learning task on the model card, and read the license before you ship it in a product. It is set up for baseRT. The card lists the license as apache-2.0.
Qwen/Qwen3-30B-A3B-Instruct-2507 converted to the baseRT .base format for local inference on Apple Silicon via BaseRT.
Access
Public
Updated Jul 30, 2026
Repo size
66.5 GB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.base66.5 GB · 100%
At a glance
- Library
- baseRT
- License
- apache-2.0
- Access
- Public
- Created
- Jul 6, 2026
- Updated
- Jul 30, 2026
- SHA
- a54e351a
Base models
Model card
From the Hugging Face model README
Qwen3-30B-A3B-Instruct-2507 (baseRT .base)
Qwen/Qwen3-30B-A3B-Instruct-2507
converted to the baseRT .base format for local inference on Apple Silicon
via BaseRT.
A 30B-parameter Mixture-of-Experts model (128 experts, 8 active per token; ~3B active params).
Files
| File | Quant | Size | Notes |
|---|---|---|---|
Qwen3-30B-A3B-Instruct-2507-Q4.base | Q4 (experts) | ~17 GB | Recommended; runs on 24–32 GB Apple Silicon |
A Q8 build (~31 GB, for machines with a larger working-set budget) is planned.
Usage
basert chat basecompute/Qwen3-30B-A3B-Instruct-2507
basert serve basecompute/Qwen3-30B-A3B-Instruct-2507
Requires Apple Silicon (M1 or later) and macOS 14+.
- Library
- baseRT
- License
- apache-2.0
- Created
- Jul 6, 2026
- Updated
- Jul 30, 2026