Managed Qwen setup inside Mallo
Install Qwen3-ASR-1.7B-8bit from the Models page with a managed runtime path instead of a manual Python setup.
Qwen setup now happens inside Mallo. No terminal-first flow. No asking regular users to understand Python environments before they can try a speech model.
What Mallo Installs For You
When you install Qwen from the Models page, Mallo now handles the full bootstrap path:
- a managed Apple Silicon Python runtime
- the app-owned virtual environment
mlx-audioand runtime dependencies- the fixed Qwen model download and cache path
Everything lives under Mallo's own Application Support directory.
Why This Feels Better
The old path worked for technical users, but it still felt like a developer setup. That created too much friction for a normal Mac app.
The new path is much closer to the expected experience:
- Pick Qwen.
- Click install.
- Wait for progress.
- Start dictating.
What It Unlocks
This is the change that makes Qwen realistic for everyday use. If you want strong Korean, mixed-language speech, or a local multilingual model that feels premium, Mallo can now get you there without external tooling.
Related updates
Keep reading
Qwen is now a first-class model in Mallo
Qwen now lives in the main model list with cleaner setup guidance, better ordering, and a more native Mallo experience.
Mallo 1.2.0
Mallo 1.2.0 brings stronger multilingual dictation, a cleaner models experience, and managed Qwen setup on macOS.
The tray got simpler while model coverage expanded
We simplified the tray menu and broadened speech model coverage so everyday dictation feels more direct.
Shipping continuously
Follow each release as Mallo gets sharper.
We’ll keep logging multilingual dictation, model setup, licensing, and workflow improvements here.