CivArchive
    ComfyUI FastVideo FastH3 4-step LoRa (unofficial) - FastH3-4-step-Preview-1.0
    NSFW

    An unofficial ComfyUI port of the FastH3 4-Step Preview LoRA.

    Sources

    Conversion recipe (as recorded in the source metadata):

    format = pt
    source_format     = Diffusers PEFT LoRA
    target_format     = ComfyUI generic LoRA
    qkv_fusion        = block diagonal B; concat A; alpha multiplied by 3
    training_rank     = 128
    training_scale    = 1.0
    training_alpha    = 128.0
    swi_glu_mapping   = Diffusers [value;gate] -> ComfyUI [gate;value]
    base_model        = Comfy-Org/MiniMax-H3 minimax_h3_fl2va_bf16.safetensors

    Compatibility

    • ✅ I2V and T2V — working on the MiniMax H3 FL2VA model (I2V 1- and T2V 2-sample videos)

    • ⚠️ Ref2VA — partially working, but loses significant prompt adherence; an official release is recommended

    Examples are made on default minimax workflow, 4 steps, Larryvrh turbo sampler

    8ish steps making sound better in my own experience.

    Workflows are imbedded in the videos.

    Credits

    All credit for the original model and the conversion recipe belongs to the original authors:

    This conversion is an independent, unofficial effort and is not affiliated with, endorsed by, or sponsored by FastVideo, lightx2v, or ModelTC.


    Right now I know about an experimental checkpoint by Kijai, would love to hear the difference from users.
    https://huggingface.co/Kijai/MiniMax-H3-experimental/tree/main

    Description

    FAQ

    Comments (17)

    BonerSoupAug 29, 2026· 1 reaction
    CivitAI

    There's only 1 video and the WF is not embedded in it

    NoiseWriggler
    Author
    Aug 29, 2026

    There is, just drag and drop it in comfy, civitai can't read output from Video Combine node.
    Second example in moderation for some reason, but it's the same workflow just without the first_frame

    alien55888Aug 29, 2026
    CivitAI

    Oh my god, this is really amazing, thank you so much for such a great turbo!

    NoiseWriggler
    Author
    Aug 29, 2026

    Thanks. Not mine though, leave a star on https://github.com/hao-ai-lab/FastVideo

    FourBunnyAug 29, 2026
    CivitAI

    4070TI 12GB 32G,0.8MP,4STEP,10S VIDEO,250S-300S

    0.8MP,8STEP,10S, 500S

    NoiseWriggler
    Author
    Aug 29, 2026· 2 reactions

    thanks for benchmarking

    FourBunnyAug 29, 2026· 1 reaction

    @NoiseWriggler Tested on the hybrid_fl2va_ref2va_b25-49 INT model; it works just as well.

    NoiseWriggler
    Author
    Aug 29, 2026

    @FourBunny oh that's very cool, thank you

    aixartsAug 29, 2026
    CivitAI

    A quick question: is this LoRA meant to be used with the "minimax h3 fast video vsa" model or the "minimax h3 standard" model?

    NoiseWriggler
    Author
    Aug 29, 2026· 2 reactions

    with the original one, I've personally tested on "minimax_h3_fl2va_pruned_fp8_scaled"

    aixartsAug 29, 2026

    @NoiseWriggler Ok! Thanks for the quick reply.

    yuduz367Aug 29, 2026
    CivitAI

    sampler and scheduler? thx

    Pat3dxAug 29, 2026· 9 reactions
    CivitAI

    Just a quick heads-up for everyone—but a very important one.

    This LoRa completely changes the prompt understanding and the aesthetic of the H3 model.

    I’ve just run about 30 tests using various turbo LoRas, fixed seeds, 0.4 and 0.5 MP settings, different prompts, 10-second clips, T2V, and R2V.

    So, if you’re generating content with this LoRa simply because the results look good—without having done A/B testing against other turbo LoRas—

    you’re missing out on completely different, much more beautiful videos and better prompt adherence.

    DaddyWolfgangAug 29, 2026

    Bingo. A lot of loras here, especially those trained using SDXL assets, are ruining the model and people complain about it sucking. And for those loras that aren't using SDXL assets, people run them at full power. They don't understad that they need to lower or raise weights.

    LoRAs will affect the model, that's the point, but you don't want them to change things so much that they ruin the original model's quality. If the lora says it's a "realistic" lora and it only makes every characters skin look smooth and plastic like... then it's not realistic. People don't have smooth skin. Period. Get off the internet, stop looking at instagram filtered faces, and look at real people.

    And if the lora has a bunch of samples that all have same face syndrome? AVOID.

    Pat3dxAug 29, 2026

    of course,it hightly depend on the prompt, like with all model the prompt is the most important base for a good generation.

    But based on my tests, when you prompt need a particular aesthetic, vivid colors, etc...the difference between this turbo lora and other is big.

    It will not make significant difference for simple prompt, simple scene.

    velantegAug 31, 2026

    This lora ruining image, it become plastic and fried.

    LORA
    MiniMax H3

    Details

    Downloads
    460
    Platform
    CivitAI
    Platform Status
    Available
    Created
    8/29/2026
    Updated
    9/2/2026
    Deleted
    -

    Files

    FastH3_comfyui.safetensors

    Mirrors

    CivitAI (1 mirrors)