Topaz Labs releases 2x faster Mac image enhancement for Wonder, Denoise Max, and Super Focus 3
Topaz Labs says Mac users can now run Wonder, Denoise Max, Super Focus 3, and Face Recovery 3 at roughly 2x local speed, with up to 4x depending on hardware and image size. Faster local processing cuts waiting time on large still enhancement workflows.

TL;DR
- Topaz Labs says its new Mac update makes local image enhancement roughly 2x faster for Wonder, Denoise Max, Super Focus 3, and Face Recovery 3, according to topazlabs' launch post.
- The company also claims some setups can reach 4x faster processing, with the spread depending on hardware, operating system, and image size, per topazlabs' launch post and topazlabs' speed reply.
- In the official announcement, Topaz Labs' PR release ties the speedup to a new acceleration layer called Topaz NeuroStream 2.
- Topaz's earlier release notes show these bigger models were already part of a push toward local processing on Mac, with the Wonder 2 local-render rollout introducing NeuroServer and the 1.6.0 release claiming major speed gains for the same model family.
You can read the official release, compare it with Topaz's earlier NeuroServer notes, and watch [awesome_visuals' short clip](src:3|awesome_visuals' reaction clip) of the faster render in action.
Mac speedups
Topaz framed this as a Mac-only acceleration update for its large local image models. The post names Wonder, Denoise Max, Super Focus 3, and Face Recovery 3, and says performance is now nearly doubled on-device for Mac users, with higher gains on some combinations of chip, OS, and image size.
The linked PR announcement adds concrete examples: Wonder 3 on an M4 dropping from 9.76 to 5.62 seconds per megapixel, Denoise Max on an M4 dropping from 10.40 to 5.54, and Super Focus 3 on an M1 dropping from 20.37 to 10.72.
NeuroStream 2
The company attributes the release to Topaz NeuroStream 2 in the official announcement. Topaz describes it as a local acceleration layer for running its larger diffusion-based models on Mac hardware, rather than a change to the models themselves.
That detail matters because these are the slower, heavier enhancement passes in the app: upscale, denoise, refocus, face recovery. Faster turnaround changes the waiting time on the kind of big still-image jobs people usually leave simmering.
NeuroServer models
This update lands on top of a broader local-processing rollout. In Topaz Photo v1.3.0-v1.3.1, Topaz introduced NeuroServer so Wonder 2 could run fully on-device instead of staying cloud-only, with the company pitching privacy and offline workflows as the payoff.
In Topaz Photo v1.6.0, Topaz said Wonder 3, Super Focus v3, Denoise Max, and Recover Faces 3 were already seeing major speed improvements from lower model load times. The new Mac release is the cleanest version of that story yet: same class of heavy local models, now tuned specifically for Apple hardware.