MiniMax 发布图像模型 MiniMax H3-Image,并公开模型权重。该模型支持生成 1600 万像素图像,4K × 4K 图像生成耗时 10-20 秒,具备指令跟随、图像编辑和最多 16 张参考图能力。目前已通过 API、MCode 和 Hailuo AI 开放,发布周 1600 万像素以上生成享 5 折,权重托管于 Hugging Face。
官方给出速度、画质与参考图控制三项指标,并开放权重,可据此判断开源图像模型的当前水位。
It's here! Introducing MiniMax H3-Image, and we're releasing the model weights!
🔥16 million pixel. 10-20 seconds. Open weights.
Built on H3, our new image model brings together three things that matter most:
⚡ Speed: Generate stunning 4K × 4K images in 10-20 seconds.
🎨 Quality: Incredible detail and photorealism, without the typical "AI look."
🎯 Control: Strong instruction following, image editing, and support for up to 16 reference images.
We believe creators shouldn't have to choose between speed, quality, and control. And we're excited to put these capabilities into the hands of the open-source community.
Available now via our API, MCode, and Hailuo AI — with 50% off 16 million pixel+ generations during launch week!
Can't wait to see what you create. 🔥
🔗 Hugging Face: http://huggingface.co/MiniMaxAI/MiniMax-H3-Image
🔗 API: http://platform.minimax.cn/docs/guides/image-generation
🔗 MCode: http://agent.minimax.cn/
来源:MiniMax (official) · x.com