What's new

Main improvements — motion physics and accuracy in responding to prompts. The model confidently handles fights, acrobatics, and other dynamic scenes, eliminating flicker and distortions. Improved ability to interpret complex textual descriptions and abstract concepts.

• Cleaner old scans: automatically removes dust/scratches; remnants can be quickly "healed" with the brush — less manual retouching.
• More natural faces: less "plastic" look; the algorithm auto-adjusts smoothing based on face size.
• Sharpening and noise handled more carefully: more fine details with fewer halos and "jaggies"; previews are closer to the final result.
• Higher-quality upscaling: small details are plausibly redrawn for tiny/soft JPEGs and frames from messengers.
• On "heavy" shots (noise, compression) — fewer blocking/banding artifacts; fabric and skin textures are better preserved.
• Smarter auto settings: more accurate scene recognition and targeted processing — fewer misses in batch processing.

What's new:

• Support for up to 3 photos — you can do incredible things; see the instructions
• Increased resolution

What's new:

• Improved inpainting quality.
• Better handling of text in images.
• Improved understanding of image context.

The service attempts to return a result at a resolution close to the original image.

PhotoStab v5. What's new:

• Reduced price.
• The new model better understands image context.

Meet Waifu Studio v2 — the long-awaited update to our anime model that brings real magic to the world of digital art! 🎨

What's new in version 2 ✨

Waifu Studio v2 includes improvements:

• Fresh data through May 2025 📅 — the model now recognizes characters from popular social games and the latest anime titles
• Improved pose stability 💃 — characters now look more natural and anatomically correct
• More composition variety 🎭 — from classic portraits to dynamic battle scenes
• Fewer watermarks 🧹 — we worked on image cleanliness and significantly reduced unwanted artifacts

Technical improvements 🔧

The new version delivers impressive results:

• More accurate rendering of facial features and emotions
• Improved handling of clothing and accessory details
• Stable generation of background elements

Note! Data on new characters is still being refined — it will be even better in future releases!

Try creating your perfect anime character right now in our image generation service! 🚀

We present a groundbreaking update to our music generation model - now it works 10 times better! ✨

Key Features 🚀

• Text-to-Music Generation 🎶 - simply describe the desired music in text, and AI will create the track
• Automatic Lyrics 📝 - if you don't want to add words, the model will create suitable lyrics for the melody
• Sample Works 🎧 - demo tracks added for Russian and English languages

Generation quality has significantly improved - create professional music of any genre right now!

Fixed a bug that in some cases caused a black screen to appear instead of the image when using the Dream Weaver model.
The app now works correctly on all supported platforms.

Significant improvements have been made to the Dream Weaver v2 model ✨

Key changes 🔧

• Enhanced detail 🔍 - the model now generates sharper, more detailed images with better rendering of small elements
• Reduced artifacts 🛠️ - common issues with distortions and unnatural elements have been fixed
• Improved stability ⚡ - more predictable and higher-quality generation results

Try the updated model now! 🚀