August 2026
Everything that shipped this month, in one place. Each entry links to the full story where there is one.
Everything that shipped in MITO in August. New things first, then models.
New in the app
Elements
Reusable Characters, Locations and Objects, saved on the canvas. Define one once, then bring it into any shot by name with an @ mention instead of rebuilding the same references scene by scene. A Character carries its personality, outfit and voice. A Location carries its visual style and key details. An Object carries its branding and references.
Keep the same character in every shot →

Director reviews your shots
Director now inspects what it generates, says what went wrong and offers a repair. Reviews arrive as suggestions in the chat, in batches, with the evidence attached, and you accept or decline each one. Failed generations show up inside the Element they belong to rather than disappearing.
The timeline moves onto the canvas
A timeline is now an asset on the canvas with its own player, so you can see the cut without leaving the board. Double-click to open it. Inside, drag tracks to reorder them, right-click a gap to delete it, set the frame rate, and send an export back to the canvas when you are done.
Instant voice cloning
Clone a voice from a sample without leaving the task, or open the clone dialog straight from Director. The language picker is searchable now instead of a wall of flags, custom voices can be deleted from the selector, and a character can take its voice from the catalog.
Projects
Filters on the projects list, a recent-projects submenu in the project dropdown, and example projects you can duplicate and pull apart. Anything shared with you, or anything on a community card, can be recreated as your own project.
Models
Seedance 2.5
Seedance 2.5 landed with a launch-week promotion and then settled into the model list, available on free plans under fair use.
That’s August. Everything above is live in MITO now.
The AI tool for cinematic video.
Generate, direct, and publish professional videos, powered by the best AI models.