joeygambino/MiniMax-H3-comfy-native-fl2va
text-to-video model by joeygambino
Older snapshot: check the original source for current information.
From the publisher
Model card & documentation
Source previewRead the publisher’s intended use, setup instructions, evaluations and limitations. The original model card is the source of truth.
The fl2va half of the comfy-native set. If you want ref2va, it is over here — same formats, same method, different job. Stock ComfyUI 0.32+ loads these directly. No ComfyUI-GGUF, no custom loader node, no dequantise-on-the-fly. ComfyUI carries its own quantisation system (comfy/quantops.py plus the comfykitchen kernels) and these files use it, so the kernels run in the quantised domain instead of unpacking to bf16 first. fl2va is the first/last-frame variant. It has no reference rows, so voice anchoring, the identity bank and reference images do nothing on it. In exchange it does the one thing ref2va cannot: it lands on a supplied frame. Measured against the same target frame: The stock-20-step control rules out the sampler. The checkpoints genuinely differ, by about 10 dB. ref2va treats a keyframe as a suggestion; fl2va treats it as the task. It can also take a first and a last frame and plan…
Read the full model card ↗ · Preview checked 2026-10-29T03:15:44.198Z
Inside the original model card — Document outline
- MiniMax-H3 fl2va — comfy-native quantisations
- Which one is fl2va, and why you would want it
- The files
- Check this before you download 20 GB
- Which format, and an honest correction
- How these were verified
- Two formats that are NOT here, and why
- Why there are no K-quants of this model anywhere
- How these were made
- What else you need
- Honest limits
- License
Headings are captured from the source. Links open the publisher’s document, not a locally hosted copy.
Documentation belongs to its respective authors. Reported project/model license: apache-2.0. A listing is not a grant of reuse or training rights. Confirm the document’s own terms at the source.
Model overview
joeygambino/MiniMax-H3-comfy-native-fl2va is a text-to-video model repository published by joeygambino on Hugging Face. The source reports the diffusion-single-file library.
This page summarizes Hub metadata. For intended use, training data, evaluation results and limitations, consult the original model card.
Model facts
- Task
- text-to-video
- Library
- diffusion-single-file
- Recent downloads (30 days)
- 4,254
- Cumulative likes
- 2
- Hugging Face trending score
- 0
- Reported safetensors parameters
- Not reported
- Architecture
- Not reported
- License
- apache-2.0
- Access
- Not gated by Hugging Face
- Created
- 2026-08-15T09:33:48.000Z
- Last modified
- 2026-08-15T19:32:13.000Z
Compatibility and lineage
Language coverage was not reported.
Reported base models: MiniMaxAI/MiniMax-H3
Parameter count is not a RAM/VRAM requirement. Check precision, quantization, context length and runtime compatibility in the model card; no hardware or API-cost claim is inferred here.
Use and evaluate
Check license terms and access requirements first. Review the model card’s documented loading instructions, evaluations and safety limitations. Benchmark scores are not imported by this directory, and popularity is not an accuracy ranking.
No model weights are downloaded or executed by AltAPIs. Never enable remote model code without reviewing it.
Source and freshness
Source: Hugging Face Hub. Metadata observed 2026-10-21T06:45:15.276Z. Daily imports are snapshots, not real-time monitoring.
Popularity and source listings do not establish security, suitability, licensing rights or benchmark performance.
Related models
QuantStack/Wan2.2-TI2V-5B-GGUF
QuantStack/Wan2.2-T2V-A14B-GGUF