Documentation Navigation
Models & pricing / Models / MiniMax-H3-Turbo-Lora
MiniMax-H3-Turbo-Lora Overview
This model's data has not been fully verified and benchmarks may be missing.
MiniMax-H3-Turbo-Lora is a video-generation AI model created by larryvrh, featuring undisclosed parameters and a context window of unknown.
ComfyUI-MiniMax-H3-Turbo is a model that accelerates the MiniMax-H3 model, allowing it to run in as few as 4 sampling steps. It is designed for joint video and synchronized audio generation and is compatible with the official H3 workflow. The model has been improved with a static-frame enhancement and is capable of producing high-quality results with reduced motion-smear and trailing ghosting.
Model Lineage & Specification
minimax-h3-turbo-lora is larryvrh's primary release in the current family.
minimax-h3-turbo-loraComparable models
| Feature | MiniMax-H3-Turbo-Lora | Wan Streamer v0.3 | Wan-Dancer | Sora 2 |
|---|---|---|---|---|
| Description | ComfyUI-MiniMax-H3-Turbo is a model that accelerates the MiniMax-H3 model, allow... | Wan Streamer v0.3 learns video as a persistent world plus the events that unfold... | Wan-Dancer is a hierarchical framework for minute-scale coherent music-to-dance ... | Flagship video generation with synced audio... |
| API Identifier | minimax-h3-turbo-lora | wan-streamer-0-3 | wan-dancer | sora-2 |
| Parameters | — | — | 14B | Unknown |
| Context Window | unknown | unknown | unknown | Unknown |
| License / Type | open weights | research preview | open weights | closed source |
MiniMax-H3-Turbo-Lora
ComfyUI-MiniMax-H3-Turbo is a model that accelerates the MiniMax-H3 model, allow...
minimax-h3-turbo-loraWan Streamer v0.3
Wan Streamer v0.3 learns video as a persistent world plus the events that unfold...
wan-streamer-0-3Wan-Dancer
Wan-Dancer is a hierarchical framework for minute-scale coherent music-to-dance ...
wan-dancerSora 2
Flagship video generation with synced audio...
sora-2ComfyUI-MiniMax-H3-Turbo is a model that accelerates the MiniMax-H3 model, allowing it to run in as few as 4 sampling steps. It is designed for joint video and synchronized audio generation and is compatible with the official H3 workflow. The model has been improved with a static-frame enhancement and is capable of producing high-quality results with reduced motion-smear and trailing ghosting.