Open-Source ComfyUI Prompt Studio Writes Prompts for Multiple Media Models at Once
tngklp · reddit · 2026-09-25
- A fork of duckyshell's ComfyUI-MiniMaxH3-Prompt-Writer, generalized from a single target into a multi-target prompt-writing workspace
- Pick a generation target, write a plain-language brief, optionally add media references, and get an editable prompt in the target's expected format; it's a UI extension, not a workflow node — it writes text only
- Supported targets: MiniMax H3 (video: T2VA/I2VA/FL2VA/L2VA/Ref2VA), Qwen Image 2.1 (T2I/edit), Krea 2 (T2I), MiniMax Music 3 (caption/lyrics)
- Standalone Windows version available; repo open-sourced on GitHub
Related event: ComfyUI's Prompt Studio Generates Prompts for Multiple Models(2 posts)→
More from Multimodal
- Opus 5.5 Turns IKEA Manuals into 3D Narrated Videos via Code — deedydas · 2026-09-25
- Testing fighting-motion video generation with MiniMax Singularity — jambonking · 2026-09-25
- Casa Toscana 3D town to sell source code with living engine and web MCP — jasonkneen · 2026-09-25
- fal opens GenMedia conference in SF, unveils H3 Max model family and fal Agent — gorkem · 2026-09-25
- Video character swap with MiniMax H3 takes 3 hours for 10-minute clips — Easy_Werewolf7903 · 2026-09-25
- Reddit: hunting the fastest model for in-video character replacement on RTX 6000 Pro — Easy_Werewolf7903 · 2026-09-25