Users Test MiniMax H3 Image Model Locally with Strong Results

Users report running MiniMax H3 image model locally on consumer GPUs: an AMD card generated a 4MP image in 25 seconds, while an RTX 4090 produced a 16MP image in about 22 seconds with 6 steps in ComfyUI.

2026-11-12 ~ 2026-11-12 · 2 related posts

• Running MiniMax-H3-Image locally: 16MP image in 6 steps, ~22s on a 4090 — metover · 2026-11-12
• MiniMax H3 image gen runs locally on AMD RX 9070, 25-30s per 4MP image — omglionheaded · 2026-11-12