# [Minimax-H3-Turbo](https://github.com/ModelTC/Minimax-H3-Turbo) Minimax-H3-Turbo provides MiniMax-H3 Turbo LoRA checkpoints, plus Diffusers batch inference and ComfyUI workflows. - **Online Studio:** [x2v.light-ai.top](https://x2v.light-ai.top/) - **API Documentation:** [x2v.light-ai.top/api-docs](https://x2v.light-ai.top/api-docs) - **Model Weights:** [Hugging Face](https://huggingface.co/lightx2v/Minimax-h3-Turbo) ## 1. Model specs
Model Tasks Training
resolution		 Training shifts
(video / audio)		 Distillation
steps (NFE)		 Recommended inference
steps (NFE)
FL2VA Turbo 4-step v0.1
Diffusers · ComfyUI 		FL2VA / T2VA 544p
mixed aspect ratio		 12 / 3 4 4
FL2VA Turbo 8-step v1.0
Diffusers · ComfyUI 		FL2VA / T2VA 544p
mixed aspect ratio		 12 / 3 8 8 / 4
FL2VA Turbo 4-step v1.0 768p
Diffusers · ComfyUI 		FL2VA / T2VA 768p
1344×768		 6 / 3 4 4
FL2VA Turbo 8-step v1.0 768p
Diffusers · ComfyUI 		FL2VA / T2VA 768p
1344×768		 6 / 3 8 8
Ref2VA Turbo 4-step v0.1
Diffusers · ComfyUI 		Ref2VA 544p
mixed aspect ratio		 12 / 3 4 4
### Note on shift For `NFE = N`, define the N transformer evaluation points on the unshifted grid as `q_i = (N - i) / N`, where `i = 0, 1, ..., N - 1`. For example, with `NFE = 4`, `video shift = 12`, and `audio shift = 3`, the shared grid is `q = [1, 0.75, 0.5, 0.25]`, giving video sigma `[1, 0.9730, 0.9231, 0.8000] -> 0` and audio sigma `[1, 0.9000, 0.7500, 0.5000] -> 0`; each list therefore uses exactly four NFEs. ### Note on reference-image resizing The three reference-image resizing policies used by our workflows are based on the `ref_image_size` implementation described in [ComfyUI's MiniMax H3 R2V reference-image sizing guidance](https://docs.comfy.org/tutorials/video/minimax/minimax-h3#prompting-tips-3):
Mode Behavior Scale factor
before 32-pixel rounding
match Matches the reference pixel area to the target canvas while preserving the reference aspect ratio. It never upscales a smaller reference. min(1, sqrt(target_area / ref_area))
max Preserves the reference aspect ratio and only scales down references whose short edge exceeds 2048 pixels. min(1, 2048 / ref_short_edge)
diffusers Preserves the reference aspect ratio and forces the short edge to 2048 pixels, matching the original Diffusers behavior. 2048 / ref_short_edge
All three policies keep the reference aspect ratio, use the H3 resolution grid (dimensions rounded to multiples of 32), and avoid cropping the reference content. In our distillation training, we use `match`, so the reference-image pixel budget follows the target training resolution. The Ref2VA inference entry point in this repository exposes the same three policies through `--reference-resize-mode` and defaults to `match`. Passing `--reference-resize-mode diffusers` restores the original Diffusers behavior (the fixed 2048-pixel short edge). For our distilled models, **we recommend selecting `match`** so inference uses the same resizing policy as training. ## 2. Online Studio and API ### Online App Try the MiniMax-H3 Turbo LoRA directly in [LightX2V Studio](https://x2v.light-ai.top/). The Studio currently uses the [FL2VA Turbo 8-step v1.0 768p LoRA](https://huggingface.co/lightx2v/Minimax-h3-Turbo/blob/main/minimax_h3_fl2v_turbo_8step_v1.0_768p_bf16.safetensors). - **Online Studio:** [https://x2v.light-ai.top/](https://x2v.light-ai.top/) - **Current model:** FL2VA Turbo 8-step v1.0 768p - **Recommended settings:** 8 NFE, video shift 6, audio shift 3 > The model version deployed in the Studio may be updated over time. ### Studio Preview ### Online API Integrate MiniMax-H3 Turbo into your application through the LightX2V API: - **API Documentation:** [https://x2v.light-ai.top/api-docs](https://x2v.light-ai.top/api-docs) ## 3. Diffusers setup and inference See [DIFFUSERS_SETUP_AND_INFERENCE.md](DIFFUSERS_SETUP_AND_INFERENCE.md) for environment setup, checkpoint downloads, test JSON files, and single- or multi-GPU inference commands. ## 4. ComfyUI inference See [COMFYUI_SETUP_AND_INFERENCE.md](COMFYUI_SETUP_AND_INFERENCE.md) for ComfyUI requirements, model installation, inputs, prompts, and run instructions. ### Example workflows Ready-to-import graphs are in [example_workflows](example_workflows/). The T2VA and I2VA graphs default to **FL2VA Turbo 8-step v1.0**; the Ref2VA graph uses **Ref2VA Turbo 4-step v0.1**. | Workflow | Task | Default resolution | |---|---|---| | [video_minimax_h3_t2v_lightx2v_turbo.json](example_workflows/video_minimax_h3_t2v_lightx2v_turbo.json) | T2VA (text-to-video + audio) | 960×544 (`16:9`, `0.5` MP) | | [video_minimax_h3_i2v_lightx2v_turbo.json](example_workflows/video_minimax_h3_i2v_lightx2v_turbo.json) | I2VA / FL2VA (image-to-video + audio) | 864×480 (`16:9`, `0.4` MP) | | [video_minimax_h3_ref2v_lightx2v_turbo.json](example_workflows/video_minimax_h3_ref2v_lightx2v_turbo.json) | Ref2VA (reference-to-video + audio) | 960×544 (`16:9`, `0.5` MP) | All graphs wrap the same MiniMax-H3 subgraph. T2VA leaves `first_frame` / `last_frame` unconnected; I2VA connects a `LoadImage` to `first_frame`, with `last_frame` optional for first/last-frame interpolation. Ref2VA connects one or more reference images through the reference-input branch. For detailed workflow inputs and execution steps, see [COMFYUI_SETUP_AND_INFERENCE.md](COMFYUI_SETUP_AND_INFERENCE.md). ## 5. Roadmap 1. Improve the visual quality and consistency of Ref2VA and FL2VA Turbo. ## 6. Acknowledgements Some Ref2VA test cases and reference assets are adapted from public showcases on the [Hailuo website](https://hailuoai.video/) and from the [MiniMax-H3 discussion on Hugging Face](https://huggingface.co/lightx2v/Minimax-h3-Turbo/discussions/29). We thank the community contributors for sharing their test assets and prompts. Special thanks to the [MiniMax-AI/MiniMax-H3](https://github.com/MiniMax-AI/MiniMax-H3) project and the MiniMax team for open-sourcing the MiniMax-H3 model.