Sign In

Text to image: MiniMax H3-Image 4K in 6 steps (ComfyUI)

Updated: Nov 21, 2026

comfyuitext2imgh3minimax

Download

1 variant available

Config Other

AIKSK_H3_Image____.json

85.97 KB

Verified:

Type
Workflows
Stats

40

Reviews
Published

Nov 21, 2026

Base Model
Hash
AutoV23F842C30FA

The FLUX.1 [dev] Model is licensed by Black Forest Labs Inc. under the FLUX.1 [dev] Non-Commercial License. Copyright Black Forest Labs Inc.

IN NO EVENT SHALL BLACK FOREST LABS INC. BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH USE OF THIS MODEL.

十六图参考编辑派对-3.png

MiniMax H3-Image text-to-image at native 4K (~16.8MP) in six sampling steps. Six-node graph: prompt in, 6-step sampling, 4K image out. Model dependencies are not bundled (minimax_h3_image_pruned_int8_convrot, minimax_h3_image_vae_fp16, qwen3.5_9b_w4a8_gptq, qwen3.8_27b_w4a8).

Episode: https://youtu.be/xz45b-opMHU

Chinese original: https://www.bilibili.com/video/BV1JaWm6FEiC/

Try online: https://www.runninghub.ai/zh-cn/post/2120808839725920258/?inviteCode=rh-v1111

Known limits: 6 steps is fast but hard prompts still need several rolls; 4K output makes 4K-sized mistakes visible.