Mallo 1.2.14
Mallo 1.2.14 makes Qwen setup more reliable, fixes the Models-page install path, and gives slow downloads more room to finish cleanly.
Mallo 1.2.14 is a Qwen reliability patch. It fixes the in-app install path from the Models page and makes large model downloads less eager to fail on slow connections.
Release Highlights
- Models-page Qwen install now completes more reliably instead of breaking partway through setup.
- Slow downloads get more room to recover before Mallo gives up on a Qwen install.
- Qwen setup is more predictable overall so fewer installs end in a dead end before the model is ready.
What Changed In Real Use
The most important fix is the one regular users actually feel: if you open Models, pick Qwen, and click install, Mallo now continues through runtime setup, package install, and model download more reliably.
This patch also smooths out setup reliability behind the scenes, so Qwen installation behaves more consistently from machine to machine.
Why This Patch Matters
Qwen is one of the most important model paths in Mallo, so install reliability matters as much as dictation quality. A local model only feels first-class if setup is boring, predictable, and resilient when the network is uneven.
This patch is about making that promise real. Fewer install dead ends, and a more forgiving download path for a large model.
Related updates
Keep reading
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 starts faster and fullscreen dictation feels smoother
Mallo now gets Qwen ready much faster on cold start, improves the default model path in onboarding, and behaves better over fullscreen apps.
Mallo 1.2.1
Mallo 1.2.1 improves HUD control, stabilizes repeat Qwen dictation attempts, and adds cleaner synthetic dictation testing paths.
Shipping continuously
Follow each release as Mallo gets sharper.
We’ll keep logging multilingual dictation, model setup, licensing, and workflow improvements here.