MiniMax H3 is starting to become a lot more interesting once you move beyond the base model. There are already LoRAs and modified versions that can change how fast it generates, how characters move, how the camera behaves, how physical interactions look, and even the overall visual style of the output.
Some of these are built for very specific jobs, while others completely change the way you can use H3 in a workflow. You will find everything from 4-step acceleration and audio-video optimization to anime line art, Wushu action, realistic people, spatial physics, Krea 2-inspired styling, and more.
The resulting videos mentioned below are highly compressed. For actual quality please refer to their Hugging face repositories. Below, we have collected the MiniMax H3 LoRAs and variants worth checking out, along with what each one is actually useful for.
Table of Contents
Instructions to do the setup:
(a) Make sure you use the Minimax H3 base model with workflow. Then, download any of the lora models from its hugging face repository and save it into your ComfyUI/models/loras folder (for Comfy users) or into specfic relevant folder if using other WebUIs like Wan2GP, ForgeUI etc.
(b) The recommended workflows and settings are available in the respective repositories. However, if a workflow is not provided, you can manually integrate the LoRA into the standard MiniMax H3 workflow.
Start with the basic MiniMax H3 workflow and add a Load LoRA node. Connect the LoRA loader after the Load Diffusion Model node or after the UNet Loader if you are using a GGUF modeland before the KSampler node. Then, select and load the corresponding downloaded LoRA model in the Load LoRA node. Once connected correctly, the LoRA will be applied to the model pipeline during generation, allowing you to use its specific style or effects with the MiniMax H3 workflow.
MiniMax-H3 Acc LoRA
Alibaba's MiniMax-H3 Acc LoRA use Parallel Decoding Distillation (PDD) to speed up MiniMax-H3 image and video generation, allowing high-quality results with just a few inference steps. The collection includes acceleration LoRAs for both FL2VA and Ref2VA workflows.
Download Minimax H3 Acc Lora model from its hugging face repository
MiniMax H3 Equirectangular 360° LoRA
MiniMax H3 Equirectangular 360° LoRA is designed to generate immersive 360° equirectangular videos using MiniMax H3. It creates a full spherical view of a single environment, with the scene wrapping completely around the viewer in every direction. The horizon is positioned around the vertical center of the frame, making the output suitable for 360° and immersive viewing experiences. The videos are generated at 24 FPS and can also include MiniMax H3's native audio capabilities, allowing both the visuals and audio to be created as part of the experience.
Download Minimax H3 Equirectangular 360 Lora model from its hugging face repository
MiniMax H3 FL2V 8-Step Motion Enhancer LoRA
MiniMax H3 FL2V 8-Step Motion Enhancer is an 8-step LoRA designed for the MiniMax-H3 FL2VS video model. It is built on top of the vanilla LightX2V 8-step LoRA and focuses on improving motion generation for human motion content.
The LoRA aims to produce more accurate male and female anatomy while better understanding motion patterns associated with close scenes. This helps improve animation quality and create smoother, more natural-looking movements in supported generations.
Download this FL2V 8-Step Motion Enhancer Lora model from its hugging face repository
Minimax H3 Semantic Bridge Adatper
This is not a lora but you can use it as the adapter model with your basic Minimax H3 workflows. It helps in enhancing prompt adherence in human anatomy etc where the base model fails.
Minimax H3 World Keyboard-Driven LoRA
Minimax H3 World Keyboard-Driven LoRA is designed to create interactive, action-controlled videos using an initial frame and keyboard inputs. By responding to keyboard controls, the LoRA can generate coordinated character actions and camera movements based on the user's input. This makes it possible to guide how the scene evolves, giving users more direct control over movement and interactions during video generation. It is particularly useful for creating dynamic, controllable scenes where both the character and camera need to move together in response to user actions.
Download World Keyboard-Driven LoRA model from its hugging face repository.
Minimax H3 Motion Continuity Repair LoRA
Minimax H3 Motion Continuity Repair LoRA is designed to improve motion consistency and continuity in dynamic video scenes generated with MiniMax H3. It can be useful for action-heavy scenarios such as running, sports, dancing, acrobatics, character interactions, combat, and weapon movements. The LoRA focuses on helping movements flow more naturally and consistently, making it particularly useful for scenes where characters or objects perform fast and complex actions.
Download Motion Continuity Repair LoRA model from its hugging face repository
MiniMax H3 VR180 Stereoscopic Side-by-Side LoRA
MiniMax H3 VR180 Stereoscopic Side-by-Side LoRA is designed to generate immersive VR180 stereoscopic videos using a side-by-side format. Each frame contains two separate views: the left-eye view on the left side and the right-eye view on the right side, creating the stereoscopic depth needed for VR viewing. Each view represents a 180° x 180° half-equirectangular hemisphere, allowing the scene to provide an immersive 180-degree perspective. The LoRA generates video at 24 FPS and supports MiniMax H3's native audio, making it suitable for creating complete VR180 video experiences with both stereoscopic visuals and synchronized audio.
Download VR180 Stereoscopic Side-by-Side LoRA model from its hugging face repository.
MiniMax H3 Human Motion LoRA
MiniMax H3 Human Motion LoRA is designed to improve the quality of human movement in MiniMax H3 text-to-video (T2V) and image-to-video (I2V) generation. It focuses on creating smoother and more natural-looking body movements, helping improve motion quality when generating videos featuring people and character actions. For best results, the recommended settings include a 720×1280 resolution, a LoRA weight between 0.4 and 0.8, and approximately 15–30 sampling steps.
Download MiniMax H3 Human Motion LoRA model from its hugging face repository.
MiniMax H3 Influencer LoRA
MiniMax H3 Influencer LoRA is designed to create videos with the polished, trendy visual aesthetics commonly seen on platforms like Instagram and TikTok. It supports both text-to-video (T2V) and image-to-video (I2V) generation, making it useful for creating influencer-style content, social media videos, and visually engaging vertical content. For recommended results, use a resolution of 720×1280, a LoRA weight between 0.75 and 0.8, and around 15–30 sampling steps.
Download MiniMax H3 Influencer LoRA model from its hugging face repository.
MiniMax H3 Facial Realism CloseUp LoRA
MiniMax H3 Facial Realism CloseUp LoRA is designed to improve facial realism in close-up, human-focused video generation. It was trained on a high-quality portrait dataset featuring detailed human faces, with a focus on capturing subtle facial features, natural expressions, and small micro-expressions. This makes it particularly useful for generating close-up shots where realistic facial detail and expressive character performance are important.
Download MiniMax H3 Facial Realism CloseUp LoRA model from its hugging face repository.
MiniMax H3 VH5Tape VHS LoRA
MiniMax H3 VH5Tape VHS LoRA is designed to give generated footage the distinctive look of video recorded from 1980s broadcast television onto an aging VHS tape. It recreates classic analog imperfections such as soft and smeared details, chroma color bleeding, tracking noise, and head-switching bands around the edges of the frame. This makes it useful for creating retro footage with an authentic, degraded VHS aesthetic.
Download MiniMax H3 VH5Tape VHS LoRA model from its hugging face repository.
FastH3 Dense 4-Step v1 LoRA
FastH3 Dense 4-Step v1 LoRA is a ComfyUI-optimized, pruned version of the FastH3-Dense-4-step-v1 LoRA for the MiniMax H3 video model. It is based on the officially released FastH3 4-step preview LoRA and uses the dense version, which does not rely on VSA. The LoRA is designed for fast video generation using only a small number of sampling steps. The official recommended settings are 4 steps, CFG 1, Euler/Simple sampling, 12 video shift, and 3 audio shift. While 4 steps can produce acceptable video quality, testing suggests that the audio quality may suffer at this setting. For better audio results, 8 steps are recommended. However, at that point, other 8-step LoRAs such as LightX2V and Alibaba's versions may offer a more attractive balance of speed and quality.
Download FastH3 Dense 4-Step v1 LoRA model from its hugging face repository.
Minimax H3 STUDIO 1939 Animation LoRA
MiniMax H3 Studio 1939 Animation LoRA gives your generations the look of classic 1930s animation. It transforms text prompts into footage with hand-painted gouache backgrounds, visible brushwork, warm celluloid-style colors, and traditional character animation aesthetics.
Download MiniMax H3 Studio 1939 Animation LoRA model from its hugging face repository
MiniMax-H3 with Z-Image-native
MiniMax-H3 plus Z-Image Native adapts Z-Image’s spatial-attention profile to MiniMax-H3, producing richer textures and a more natural, gritty look while reducing digital over-sharpening artifacts. It’s formatted for native ComfyUI integration and also includes VRAM-friendly 4-bit versions for more efficient local use.
Download MiniMax-H3 plus Z-Image Native model from its hugging face repository.
MiniMax H3 Turbo Sparse Linear Attention Lora
MiniMax H3 Turbo SLA LoRA uses Sparse–Linear Attention with an 85% attention sparsity ratio to make inference more efficient. It delivers around 2.5× faster inference on an NVIDIA RTX 5090 while maintaining competitive visual quality.
Download MiniMax H3 Turbo SLA LoRA model from its hugging face repository.
MiniMax H3 Turbo 4-Step V1.1 LoRA
LightX2V’s MiniMax H3 Turbo 4-Step (v1.1) is an accelerated version of MiniMax H3 designed to generate results in just 4 inference steps, making the model significantly faster while keeping visual quality competitive.
Download MiniMax H3 Turbo 4-Step (v1.1) lora model from its hugging face repository.
MiniMax-H3 Looping Sketch Anime Lora
MiniMax-H3 Looping Sketch Anime LoRA is designed for creating looping animations with a hand-drawn anime sketch aesthetic. It can turn text prompts into short, repeating sequences with expressive linework and a rough, illustrated look, making it well suited for animated sketches, character movements, and other seamless looping scenes.
Download MiniMax-H3 Looping Sketch Anime LoRA model from its hugging face repository
Minimax H3 plus Krea 2 Style Lora
MiniMax H3 cum Krea 2 Style LoRA is an experimental style-shifting LoRA that brings the visual aesthetic of Krea 2 into the MiniMax-H3 video generation model. It is designed primarily for Text-to-Video (T2V) workflows, helping generate videos with a Krea 2-inspired visual style while retaining MiniMax-H3's video generation capabilities.
Download MiniMax H3 + Krea 2 Style LoRA model from its hugging face repository
MiniMax H3 RAVEN Streaming LoRA
MiniMax H3 RAVEN Streaming LoRA integrates the RAVEN framework for real-time, streaming video generation and long-horizon extrapolation. It is designed to maintain video quality over longer sequences, helping reduce the gradual quality degradation and consistency issues that can occur as a video progresses.
Download MiniMax H3 RAVEN Streaming LoRA model from its hugging face repository
MiniMax-H3 Prompt Rewriter LoRA 8B adapter
LightX2V's MiniMax-H3 Prompt Rewriter LoRA 8B Adapter helps turn short, simple ideas into detailed prompts that are better suited for MiniMax-H3’s audio and video generation. It works with both text-based and keyframe-conditioned inputs and supports T2VA, T2AV, I2VA, I2AV, L2VA, L2AV, and FL2VA workflows, making it easier to get more descriptive and production-ready results without writing complex prompts yourself.
Download MiniMax-H3 Prompt Rewriter LoRA model from its hugging face repository
Wan2.2 plus Minimax H3 Motion Lora
Wan2.2-MiniMax H3 Motion LoRA is designed to improve motion and movement in videos generated with MiniMax H3. It combines motion-related capabilities from Wan2.2 with MiniMax H3, helping produce more dynamic and natural-looking movement while keeping the core visual style of the H3 model.
Download Wan2.2-MiniMax H3 Motion LoRA model from its hugging face repository
Minimax H3 Wushu Action LoRA
MiniMax H3 Wushu Action LoRA adds a focus on martial arts and kung fu movements to MiniMax H3. It helps generate videos featuring human physical actions such as punches, kicks, staff techniques, and other dynamic Wushu-style movements, making it useful for creating action-heavy martial arts scenes.
Download MiniMax H3 Wushu Action LoRA model from its hugging face repository
Minimax H3 Camera Motion LoRA
MiniMax H3 Camera Motion LoRA adds more precise cinematic camera control to MiniMax H3. It helps guide different types of camera movements, including push, pull, pan, tilt, circular moves, and handheld motion, making it easier to create videos with more intentional and dynamic camera work.
Download MiniMax H3 Camera Motion LoRA model from its hugging face repository
Minimax H3 Spatial Physics Lora
MiniMax H3 Spatial Physics LoRA helps MiniMax H3 better understand spatial relationships and physical interactions between objects. It is designed to improve how the model handles actions such as collisions, stacking, falling, and occlusion, making object movement and interactions in generated videos feel more physically consistent and believable.
Download MiniMax H3 Spatial Physics LoRA model from its hugging face repository
Minimax H3 Latent Upscaler
MiniMax H3 Latent Upscaler is designed to increase the spatial resolution of MiniMax H3 video outputs while preserving the original temporal dimension. Trained on around 80,000 paired video and image samples, it works directly with MiniMax H3’s 24-channel VAE latents to upscale the height and width of generated content while maintaining its motion and temporal structure.
Download MiniMax H3 Latent Upscaler model from its hugging face repository
MiniMax-H3 Anime Video Line Art Colorization lora
MiniMax-H3 Anime Video Line Art Colorization LoRA is designed to turn anime-style line art videos into detailed, colored visuals. It takes line art as the input and helps MiniMax H3 generate polished anime scenes with richer colors and refined visual details while preserving the original line-art structure.
Download MiniMax-H3 Anime Video Line Art Colorization LoRA model from its hugging face repository
10 Eros Minimax H3
You obviously know 10 eros' models are famous for what. 10 Eros MiniMax H3 is a custom MiniMax H3 base model that combines learned patterns from LTX 2.3 and Wan 2.2 video diffusion models. It is intended to work best with image-to-video (I2V) or reference-based workflows without additional LoRAs. For the best results, the model recommends following the official MiniMax H3 prompting style when writing prompts.
Download 10 Eros MiniMax H3 base model from its hugging face repository
Minimax H3 Realism People LoRA
MiniMax H3 Realism People LoRA is designed to make human characters look more realistic and natural in generated videos. It focuses on details such as convincing facial features, natural skin texture, realistic expressions and gestures, along with film-style lighting and documentary-inspired camera movement.
Download MiniMax H3 Realism People LoRA model from its hugging face repository
MiniMax H3 Acceleration LoRA
MiniMax H3 Acceleration LoRA is designed to speed up MiniMax H3 video generation for both Text-to-Video (T2V) and Image-to-Video (I2V) workflows. It targets H3’s joint audio-video diffusion process and uses a fixed 4-step Euler inference setup, allowing faster generation with fewer sampling steps.
Download MiniMax H3 Acceleration LoRA model from its hugging face repository
MiniMax-H3 Audio-Video acceleration LoRA
MiniMax-H3 Audio-Video Acceleration LoRA speeds up the model's audio-video generation by reducing the standard 20-NFE process to a fixed 8-NFE Euler workflow. It is designed to maintain MiniMax-H3’s native stereo audio, dialogue, and synchronized sound effects, while requiring fewer inference steps for faster generation.
Download MiniMax-H3 Audio-Video Acceleration LoRA model from Hugging face repository.
