What it does

Sisif turns a text prompt, or an uploaded photo plus a prompt, into a short video clip. A standard clip is five seconds of portrait MP4 at 540x960 (720x1280 on an ultra tier), and most clips are ready in two to five minutes. Longer briefs can be planned as a storyboard-based story of up to 60 seconds, with the price shown before you pay. The site's stated terms are no watermark at any price, credits that never expire, an automatic credit return when a render fails, and commercial use included. It says clips are rendered by an open-weights video model on GPUs the company runs, with no third-party API involved. Clips can be made on the website, through a REST API with webhooks, or by an agent through an MCP server. Agents can pay per clip over Stripe's machine payments protocol, and an n8n template is mentioned. Sisif says it has no free tier and no free trial credits, and a new account starts with zero credits. Signing up, using prompt suggestions and planning storyboards are free.

Capabilities

Text-to-video clips of five seconds in portrait MP4
Animates an uploaded photo from a prompt
Storyboard planning for stories up to 60 seconds, priced before payment
No watermark, non-expiring credits and credits returned for failed renders
REST API, webhooks, n8n template and an MCP server for agents
Agents can pay per clip over MPP

When to use it

Making short vertical clips for TikTok, Reels, Shorts or ads
Animating a photo, such as an old family photo or a product shot
Producing logo intros or faceless-channel videos
Letting an AI agent order and pay for single clips as it needs them

Alternatives to explore

Kling
Runway
Pika