HyperKut Studio: open-source multi-shot AI video studio built on ComfyUI
Main_Creme9190 · reddit · 2026-10-30
A developer submitted HyperKut Studio to the Comfy Open Source Higgsfield challenge — an MIT-licensed, local-first studio that turns a reference edit, a text idea, or your own music track into a short multi-shot AI video.
Three modes: Remix reads cuts, framing and rhythm from an edit and redraws every shot with your subject; Text has Gemma 4 write the shot list, MiniMax H3 render shots, and YuE2 create the soundtrack; Music places cuts on the beats.
Editable by design: regenerate single shots, swap takes on a timeline, trim and reorder after generation with no re-render.
Real numbers (RTX 4090 24GB, model loading included): 5 min median for Draft, 10.6 min for Max, 11.5 min for Standard; a 28-shot fashion Remix took 21 min. Every mode is an open ComfyUI workflow built from the author's Majoor VideoTools nodes; the studio is FastAPI + React talking to ComfyUI via the Comfy SDK. Windows + NVIDIA only for now.
Related event: HyperKut Studio: Open-Source Multi-Shot AI Video Workbench on ComfyUI(2 posts)→
More from coding & agent
- LangChain ships managed-deep-agents 0.10 with async subagents and browser use — hwchase17 · 2026-10-30
- New paper asks why teams of AI agents fail where a single agent often matches them — JoshPurtell · 2026-10-30
- His app is live with users — he has no idea where it's hosted: "Codex" — gabe_ragland · 2026-10-30
- Superconductor launches: multiplayer AI workspace putting teams and coding agents in one session — sergeykarayev · 2026-10-30
- YC F26 Startup Hue Launches Simulated Environments for Testing AI Agents — ycombinator · 2026-10-30
- CMU Releases SwarmChat, an Open-Source Slack for AI Agent Swarms — niloofar_mire · 2026-10-30