Vaethat

Vaethat's whole pitch is restraint, and for once that's the selling point rather than a limitation. Most AI upscalers make you prompt, tune sliders, and run multiple test images to find settings that don't wreck the render. Vaethat collapses that into one decision: 2x or 4x, plus a subject context (Architecture, Exteriors, Interiors, Vegetation, People) and the optional Region and People regions. The claim is that one mode works across a whole set of renders — which is exactly what archviz studios need, because consistency across the images in a submission or a client deck matters more than squeezing the last 5% out of any single frame.Strengths: the batch Multi-Image Workspace enhances an entire project folder in parallel on one canvas with no queue, and you can apply the same mode and resolution to every image in a click. Parallel processing isn't limited to one image on Basic (1 at a time) but scales to 4 concurrent on Professional and Business, which is the real reason a studio moves off the entry tier. Region enhancement lets you lock the AI onto a focal point — a face, a fixture, a façade detail — without a full re-process, and credits scale with output pixel area (Math.ceil(sqrt(width × height) / 500)), so skipping a full re-process is a direct saving. The layered .psd export, with your original render as a pixel-aligned base layer and a region returned as its own layer, is the feature that makes this a post-production tool rather than a novelty: you keep the untouched source and blend the AI result against it in Photoshop or Affinity.Weaknesses and honest boundaries: Vaethat only enhances existing renders. It does not generate new images, does not touch video or 3D models, and is not a text-to-image design tool. Basic caps output at 4K and processes one image at a time; 64-megapixel output and concurrent processing start at Professional (€29/month). High-resolution work burns credits faster by design, and Enhance Upscale can shift small elements slightly, so you inspect before delivery. Prices exclude VAT.The closer question is where it sits among alternatives. Magnific is broader and generative — it will reimagine your render, which some artists want and many archviz clients do not. Topaz Photo AI is a general-purpose photo and video enhancer that has no archviz vocabulary at all and therefore no reason to protect straight lines or material families. Vaethat sits between them as a narrow, repeatable post step. If your workflow is V-Ray/Lumion/Corona/Twinmotion/D5 renders that need to look sharper without looking different, this is a low-friction trial; if you need one tool to cover photos, video, and generative exploration, you'll keep a second subscription alongside it.