Skip to content

Release history

Change Log

Track all the updates and improvements we've made to Kolbo.AI.

90 updates

Latest release

VIDEO FEATURES

1
  • Avatars: make an avatar of yourself, or anyone who has given consent, from one photo, a Visual DNA or a short talking video, then have it speak any script in 12 languages, Hebrew and Arabic included. Choose Omni for a full scene with real camera styles, Photo for a photo that talks in a place you pick, or Twin to put new words on your own video. Camera styles: Talking head, Podcast, Selfie and Walk and talk, across 60+ rooms and places, including cars. Use one of about 780 library voices or your own cloned voice, and add a held mic, Looks (outfits) or background removal. Find it in Video Tools, Avatars.

NEW MODELS

1
  • FLUX 3 Max: the strongest FLUX 3 video model, with sharper prompt adherence, richer visuals and native audio. Create from text, an image, first and last frames or up to 10 keyframes, 5 to 20 seconds in 720p or 1080p.

NEW TOOLS

5
  • Trends: you, in any viral trend, at app.kolbo.ai/trends. Pick a trend, add your photo (up to 3 photos of the same person, or a Visual DNA) or your own video, and render. No prompt needed.
  • Preview 4 seconds first: test any trend on your own photo or video before the full render. Results start as a 480p Draft, with a one-click Full Quality 1080p render.
  • Video recasts, Replace Character and Replace Background: put yourself into viral dance, music and action clips, or change who is in your own video, or where it happens, while keeping your real room and performance.
  • Photodump sets: turn one photo into 8 photos in a single run.
  • Reuse, favorites and My generations: reuse a past run's inputs, favorite trends, and find everything you made under My generations.

ORGANIZATION & WORKFLOW

1
  • Invite anyone to a project by email: no Kolbo account needed. They get a signup link, land on the shared project the moment they verify, and you can track, resend, or cancel invitations still pending.

VIDEO FEATURES

1
  • Use my original voice: on Elements clips built from uploaded dialogue, put your recording onto a new shot. The generated take stays.

NEW MODELS

2
  • Gemini 3.7 Flash: Google's newest Flash-tier Gemini is now in Chat and Kobi Code, with 1M context and multimodal input for text, images, video, and audio.
  • Grok 4.6: xAI's newest flagship is now in Chat and Kobi Code, with 500K context, native vision, and unified reasoning.