CreaPrompt Update: Integrates Local Qwen3-VL Prompt Enhancement
Away_Exam_4586 · reddit · 2026-07-07
The ComfyUI prompt builder node CreaPrompt received a major update featuring a built-in local LLM prompt enhancer. Powered by Qwen3-VL (defaults to Qwen3-VL-4B, auto-downloads on first run), it rewrites assembled keywords into natural language prompts tailored for modern models like Flux, Krea 2, Z-Image, Qwen-Image, and Wan. The node supports 3 image inputs and 1 video input, allowing direct prompt generation from a single image or fusing subjects, styles, lighting, and atmosphere from multiple images into a cohesive scene.
More from Multimodal
- TischEins, a large Midjourney SREF library, relaunches with new design after security outage — tisch_eins · 2026-10-26
- MOSS-Image-SFT 4-step distillation tops 16.5K downloads on Hugging Face — _akhaliq · 2026-10-26
- Claude Code can now edit videos: a 10-prompt workflow from raw footage to final cut — TawohAwa · 2026-10-26
- One-sentence 3D worlds turn cinematic with CineMap and TapNow — gizakdag · 2026-10-26
- No Camera Needed: Creator Builds an Entire World in CineMap and Directs It Into Cinema — gizakdag · 2026-10-26
- Kling vs. Seedance: the Chinese AI video model rivalry — DavidmComfort · 2026-10-26