Turkish creator Ozan Sihay released a seven-minute one-person AI short film built with Seedance 2.0, Kling 3.0, Nano Banana 2, Runway, HeyGen, Suno, and CapCut. The film matters because it turns Seedance’s weak face realism into a masked-character design rule and shows the planning graph behind the finished cut.
You can watch the full film on YouTube, browse ByteDance's official Seedance 2.0 page, and compare that with Runway's pitch for node-based Workflows. Kling's official site says its 3.0 video models are built for multimodal instructions and consistency across longer sequences, which helps explain why Sihay split visible-face shots into a different toolchain altogether (Kling AI).
The film's premise is already absurd enough to stick: a neighborhood uncle's cat gets kidnapped, he opens a car trunk, pulls out a knitted tiger mask, and fights back with a superhuman slap Sihay release thread.
The more useful detail is the production choice behind it. Sihay says Seedance 2.0 struggled with realistic human faces, so he stopped treating that as a bug to hide and made masks central to the story. The tiger mask and rusty dog mask became character design, not a workaround Sihay release thread.
Sihay's thread reads like a current-gen solo filmmaker stack:
That split lines up with how the tools describe themselves. ByteDance's official Seedance 2.0 page emphasizes multimodal audio, video, image, and text inputs, while Kling AI pitches its 3.0 models around deeper multimodal control and stronger visual-audio binding.
Sihay posted a screenshot of what he said was only a small part of the Runway Workflow behind the film workflow post. Runway describes Workflows as a node-based system for chaining models, prompts, and refinements into reusable pipelines, and the screenshot looks exactly like that: image blocks, text nodes, poster frames, characters, and branching connections across a large canvas.
That image adds one concrete thing the finished short cannot show on its own. The seven-minute cut may read like a single expressive artifact, but the making of it looks much closer to systems design.
