whatmodel.sh / image / MiniMax H3-Image

MiniMax H3-Image is an open-weights image model from MiniMax, released November 2026. Its published size is 33.1B parameters. Licence: minimax-h3-image-community.

Catalogue identity notes

MiniMax announcement (2026-11-11): 'Today, we're releasing MiniMax H3-Image, an open-weight image generation model built on H3.' Open weights at MiniMaxAI/MiniMax-H3-Image (pinned 5f90473; repository created 2026-11-03, weights committed 2026-11-10) under the MiniMax-H3-Image Community License Agreement, a custom licence, not an OSI licence. Also served on MiniMax's API as MiniMax-H3-Image and MiniMax-H3-Image-Highspeed; MiniMax prices it per token, not per image, so no price is recorded. No run command is listed yet. Distinct from the MiniMax H3 video model. Arena lists it as 'MiniMax H3-Image' on its text-to-image and image-edit boards.

Model information sources

Release date ↗ · Weights licence ↗ · Parameter count ↗

Model card: a 33.1B-parameter DiT; generation and editing in one checkpoint.

67.9±2.1#19 of 147 image modelsfrom 2 benchmark results across 1 source* thinly tested, treat the rank as provisional

Score within the LMArena text-to-image comparison set only.

$ how it compares

The image models ranked either side of MiniMax H3-Image, out of 200 with a score, then more from MiniMax.

#modelscorewhereRAMreleased
19GPT Image 1.5 OpenAI70±1api $0.034/image2025-12
20Qwen-Image-2.1 Alibaba68±2local~12GB2026-09
21MiniMax H3-Image MiniMax68±2api price unknown2026-11
22Ideogram 4.5 Ideogram67±3comparison only
23Qwen-Image-2.0-Pro (2026-06-22) Alibaba66±2api $0.075/image2026-06
more from MiniMax
94MiniMax image-01 MiniMax52±5api $0.0035/image

see all image models · find the ones that fit your machine

$ what it's good at

text to image▂▅▅▃▄▄█▃▂▁▁▁ #20 of 129 * · 73% of its score · 1 test

Trails Ideogram 4.0, Qwen-Image 2.0 Pro here, though they rank lower overall.

testresultvs the others in its rundifficultycountswhy
LMArena text-to-image1197.2 ✎▁▁▃▃▆█▄▂▂▂▁▁ beats 58/78avg 1030▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 73%typical
image to image / editing▁▁·▃▂▃█▆▃▆▂▂ #17 of 65 * · 25% of its score · 1 test

Beats Nano Banana 2 Lite (Gemini 3.1 Flash-Lite Image), GPT Image 1.5 here, though they rank higher overall. Trails Grok Imagine Image Pro here, though they rank lower overall.

testresultvs the others in its rundifficultycountswhy
Arena image edit1358.9 ✎▂▁▁▃▆█▅▇█▄▃▃ beats 38/56avg 1160▇▇▇▇▇▇▇▇▇▇ 25%typical

$ where to run it

Through an API

api · price unknown · platform.minimax.io ↗

Model repository and documentation: huggingface.co/MiniMaxAI ↗

$ how the score was worked out · LMArena leaderboard dataset 98%

Each result implies a score on its own, after adjusting for how hard its test is and who ran it. The score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its counts column above; here it's added up by who ran the tests. How it works

sourceresultsshare of the scorelicence
LMArena leaderboard dataset2CC BY 4.0
prior · pull toward the middle
weighted average= 67.9

$ improve this page

Use a short GitHub form or edit the data through a pull request. Every number needs a source; here's which ones we can use.

data updated · sources and licences · json