Testing Minimax H3: Generating Consistent Video from Dual Character Sheets and Scene
JahJedi · reddit · 2026-08-13
A developer shared their workflow and test results using the Minimax H3 (ref2av) model for video generation.
- Inputs: Used two character reference sheets and a school photo as visual guides.
- Parameters: Configured at 1.5 megapixels and 30 steps, with sage attention enabled and sol + cashe disabled.
- Hardware: Rendered on an RTX 6000 Pro. The author plans to share the full prompt and workflow files if there is community interest.
Related event: MiniMax H3 Overseas Tests: Multi-Style Long Video Generation Praised(13 posts)→
More from Multimodal
- 1986 to 2026 in one session: outpaint, modernize and 4x upscale, session file included — juanpablogc · 2026-10-16
- 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