Pixio August 2025 Updates

Discover the latest features, improvements, and fixes in our August 2025 update including new AI tools, enhanced performance, and bug fixes

Alisher Farhadi's pictureAlisher Farhadiยท
ยท5 min read
Cover Image for Pixio August 2025 Updates

โœจ August 2025 Highlights: Several game-changing additions to our platform - revolutionary image generation with text mastery and professional music creation capabilities.

๐ŸŽจ QwenCraft

Based on Qwen Image. Generate and edit images with great consistency. Good coherency for text and realism. Allows mask-free editing.

๐Ÿ“ธ QwenCraft Example

๐ŸŽฏ Prompt: "man in a black hoodie with tsi logo on the chest sitting in the dark in front of the computer"

๐Ÿ“ Input Reference
Input reference
โœจ Generated Result
QwenCraft generated result

๐ŸŽต TempoCraft Pro

Added ElevenLabs music generation. Generate a song from just a few lines of prompting and also have a much more detailed and fine grained creation experience with the composition plan. Composition plan allows specific lyrics, timing, sound changes and more. Able to generate up to 5 minutes of audio.

๐ŸŽง Audio Example (prompt + composition plan):

๐Ÿ”ฅ Notable Updates

๐ŸŽ™๏ธ Text to Dialogue in VoiceCraft

Uses ElevenLabs V3. Supports cloned voices, up to 4 speakers and prolonged generations. Ideal for podcasts and for having a back and forth between AI generated voices.

๐ŸŽต Multi-speaker dialogue example:

๐Ÿ–ผ๏ธ Nano Banana Model Update in Pixio Chat

Updated model provides some of the best consistency between edits. Allows you to use natural language to edit instead of prompting. Allows up to 5 reference images to build your scene.

๐ŸŽจ Editing Example

๐Ÿ’ฌ Prompt: "Put the hoodie with TSI logo from image 2 onto the man in image 1"

๐Ÿ“ท Image 1
Base image
๐Ÿ‘• Image 2
Reference hoodie
โœจ Result
Final edited result

๐ŸŽจ Image Generation Updates

๐ŸŽฌ Video Generation Updates

๐ŸŽต Audio Updates

๐ŸŽ™๏ธ PodCraft
Based on VibeVoice. Use a snippet of your own voices and generate dialogue with up to 4 speakers.

๐ŸงŠ 3D Generation Updates

๐Ÿ—๏ธ PolyCraft
Updated to use new models for 3.0 and 1.0 Turbo. 3.0 has high poly count and some of the highest quality.

๐Ÿ› ๏ธ Bug Fixes

โš™๏ธ Model Infrastructure Upgrade: Updated all generators to use new GPT OSS 120b model

๐Ÿš€ What's Coming Next

  • WonderCraft in Audio for podcasts
  • Updated prompt optimizer