Local H3 Image-to-Video Workflow: Krea + KdenLive Tested on RTX 3090Ti
underlogic0 · reddit · 2026-08-09
A developer successfully generated a visually striking animated short on local hardware (3090Ti + 64GB RAM) using a default ComfyUI workflow integrated with Krea's H3 (i2v) model and KdenLive.
Workflow Details:
- Efficiency: By wiring in RTX Super Resolution, Apply Spectrum, and SageAttention, generation takes only 20 steps and 3-6 minutes per clip.
- Prompt Engineering: The author used Claude, referencing the H3 prompting guide, to beef up the prompts. For fully local setups, Qwen or Gemma are recommended as alternatives.
- Post-production: Five generated clips were edited together, manually cutting out janky frames to achieve results comparable to Seedance.
More from Multimodal
- Meituan LongCat's WorldCast benchmark: GPT-Image-2.5 tops at 72.8, open-source lags — meituan-longcat · 2026-10-16
- Three resources to get up to speed with FLUX 3 — ariG23498 · 2026-10-16
- Watch: Claude Opus 5.5 builds a Tesla Roadster from a single prompt — TAbrodi · 2026-10-16
- Midjourney v8.2 turns out striking indigo ink-wash art with sref 1848359068 — azed_ai · 2026-10-16
- "AI television from another dimension": robot grooms a chihuahua on air — BotLordBotflix · 2026-10-16
- 108M-parameter text-to-image denoiser Surjo-Image-Preview lands on Hugging Face — Consistent-Rule-3720 · 2026-10-16