# MiniMax H3-Image: Native 4K at 16.8MP in Six Steps - Three ComfyUI Workflows Compared

> Source: <https://dev.to/ai-ksk/minimax-h3-image-native-4k-at-168mp-in-six-steps-three-comfyui-workflows-compared-2am2>
> Published: 2026-11-21 10:09:34+00:00

**Recommendation:** Run the six-node text-to-image graph first to validate H3-Image at 4K on your hardware, then move to the 48-node edit graphs and test the 16-image branch for identity drift when references disagree.
More in the playbook "Inference and self-hosting": <https://wpnews.pro/playbooks/inference-serving>

MiniMax H3-Image: native 4K at 16.8 megapixels in six sampling steps (three ComfyUI workflows)

MiniMax's H3-Image is unusual for one reason: it does not upscale its way to a big picture. It generates natively at 4K-class resolution - about 16.8 megapixels - and it does it in six sampling steps. This episode tests that claim with three separate ComfyUI workflows, each built for a different job.

Who this is for

The three workflows, and when to use each

Text to image - `AIKSK_H3_Image_文生图.json`

The simplest chain: prompt in, 6 steps of sampling, 4K image out. Six nodes total, so you can read the whole graph in a minute. Use it when the image does not exist yet.

Online: [https://www.runninghub.ai/zh-cn/post/2120808839725920258/?inviteCode=rh-v1111](https://www.runninghub.ai/zh-cn/post/2120808839725920258/?inviteCode=rh-v1111)

Single image edit - `H3_Image_单图像编辑.json`

48 nodes. One reference image goes in with an instruction, and the model re-renders while keeping the parts you did not ask it to change. This is the workflow for "same photo, different weather / different light / fix this corner".

Online: [https://www.runninghub.ai/zh-cn/post/2120808831056293890/?inviteCode=rh-v1111](https://www.runninghub.ai/zh-cn/post/2120808831056293890/?inviteCode=rh-v1111)

16-image reference edit - `H3_Image_16图编辑.json`

Also 48 nodes, but the reference side is a bank of up to sixteen images. That is what you want for character consistency across angles, or for feeding a whole product set into one edit so the result matches all of them.

Online: [https://www.runninghub.ai/zh-cn/post/2120808834377023490/?inviteCode=rh-v1111](https://www.runninghub.ai/zh-cn/post/2120808834377023490/?inviteCode=rh-v1111)

What is inside the JSON

`qwen3.8_27b_w4a8` for the prompt path.`minimax_h3_image_pruned_int8_convrot`, `minimax_h3_image_vae_fp16`, `qwen3.5_9b_w4a8_gptq`, `qwen3vl_32b_minimax_h3_int8_convrot`.
No model weights are bundled in the repos. You bring the weights; the workflow carries the wiring, the sampling settings and the conditioning structure.

Known limits

How to try it without a local install

All three run as online workflows on RunningHub, which is what we used for the tests in the episode - same graph, no GPU required on your side:

Episode links

Where this goes next

The interesting question is not whether 4K native generation looks good in a hero shot - it does. It is whether the 16-image branch holds identity when the references disagree with each other. That is the test worth running yourself.
