底部赞助展示区
Contact us 适合多赞助商长期展示
Loading…
适合多赞助商长期展示
A community creator demonstrated local video generation with MiniMax H3 and ComfyUI on a 24GB RTX 4090, covering text-to-video, beat synchronization, and image-to-video workflows.
A creator has shared a hands-on account of producing AI videos locally with MiniMax H3 and ComfyUI. The experiments included text-to-video, cuts synchronized to music beats, and image-to-video generation. The workflow starts with rough creative ideas, which are then converted into structured prompts using MiniMax’s official H3 prompting guidance and refined iteratively.
The creator reports that a 15-second video at 20 steps takes more than 200 seconds on a consumer RTX 4090 with 24GB of VRAM. Rendering at 768P can take more than 1,000 seconds, while upscaling to 2K through the official API reportedly takes roughly six to eight minutes. Actual performance will vary with resolution, quantization, sampling settings, and model-loading overhead.
Related community documentation indicates that H3 can be used in local ComfyUI workflows and that official templates and prompt documentation lower the barrier for new users. The creator also plans to publish the workflows, scripts, and assets used in the tests, although the reported timings and qualitative results remain personal observations rather than standardized benchmarks.
Source review does not guarantee permanent accuracy. Consider the original source, observation time, and later corrections.
`nvfp4_awq` `pruned_int8_convrot` `fp8_scaled` First run: start small. Template library → Video → MiniMax H3 → Text to Video (the three `api_`-prefixed templates call the cloud API and need a key — ignore them for now). Check that the UNETLoader points at the FL2VA weight, the encoder is the NVFP4 build, and both VAEs are loaded. A sensible 12 GB starting point: 16:9, 864×480 (~0.4 MP), 5 seconds, Turbo on, Turbo steps 8 — this validates the environment rather than chasing quality. Keep the prompt to one scene, one main action, one shot, and sound. The first generation is always slow (about 34 GB of weights load first; reference: 4090 48G first run ≈ 644 s vs ≈ 425 s warm). `api_` [...] Prompt format. H3 uses an official structured prompt format (full Prompting Guidance in the official README). A practical trick: hand your Chinese/plain brief plus the official prompt guide to any LLM and let it rewrite into the structured format, then save as a template — an effective community stand
How I Started Making AI Videos with Audio Using 'MiniMax H3' you need a VRAM 24GB class GPU. That's the world of the RTX 4090 or 5090. This
MiniMax H3 for ComfyUI is here! I am getting insane results out of it on just a RTX 4090 laptop. Leaps and bounds if we can get a replacement
A rough tiering, based on what’s been demonstrated: 6GB VRAM (RTX 2060 class): workable but degraded, expect mushy detail and rougher audio, generation times around 10-15 minutes for short clips. 12-16GB VRAM (RTX 4060 class): noticeably cleaner, faster generation, still below full quality. 24GB VRAM (RTX 4090 class): enough headroom to also do LoRA training on top of generation. 32GB+ VRAM (RTX 5090, RTX 6000): the intended target, supports longer 15-second clips at higher resolution with less compromise. The pattern lines up with most modern diffusion-based video models: more VRAM buys you resolution, generation length, and speed, but low-VRAM setups aren’t dead ends anymore thanks to aggressive quantization and optimization work from the open-source community. [...] MiniMax H3 is fully open source, meaning anyone can download the weights and run generation locally instead of relying on a paid API. VRAM needs scale from roughly 6GB to 24GB+, with lower-VRAM setups producing so
The storybook film took an afternoon once the pipeline worked: six illustrations, three techniques, about 25 minutes of GPU time for the shots plus one reference-mode generation. A year ago this was a render-farm conversation. ## Workflows for download Everything in this post runs on core ComfyUI nodes — v0.30 or later, no custom node packs to install. The weights are on Hugging Face at Comfy-Org/MiniMax-H3, and MiniMax’s structured prompt guides ship in the model repo’s `docs/` folder (`VIDEO_PROMPT_WRITING_GUIDE_base_en.md` and the `_ref_` variant for reference mode). [...] Meanwhile, rendering the same freeform prompt at 5× the pixels changed nothing: 608×352 scored 29.2, 864×480 scored 30.3, 1344×768 scored 26.0. If your generations look soft, the fix is not more resolution. To be fair to the model: the atmospheric prompt asked for shallow depth of field, film grain, sea mist, and volumetric fog — and H3 rendered exactly that. It was doing its job. But the gap between hand-tuned
Very useful resources on making videos with the Minimax H3 model on an RTX 4090 (24GB) 😊 The showcases look amazing, and detailed prompts plus