Multi-modal AI · in private beta

AI for short looping video.

Generate image, video, audio, 3D, and chained workflows in one place — the latest open and frontier models, built around the 10-second loop.

Inside the beta

One workspace for every modality.

55+ models, one wallet

Sora 2, Veo 3.1, Kling 3.0, HappyHorse, FLUX.2, GPT Image 2, Ideogram V4, ElevenLabs — 12 categories, refreshed as the frontier moves. No juggling provider keys.

Built for 10-second loops

Coub's native format is the default. Native-loop generation and 130+ viral effect presets today — loop editor and one-click publish to your Coub channel are next.

Workflows, not just prompts

Chain text → image → video → audio in a single run. 12 ready-made workflows, every step's prompt editable, start from your own image.

We called it

Coub picked 10 seconds in 2012. AI caught up 14 years later.

Visit coub.com →

FAQ

AI for short looping video.

One workspace for every modality.

Quick answers.

Get in before the public launch.

AI for short looping video.

When does it launch?

What's in early access?

Will my Coub account work here?

Pricing?

Get in before the public launch.