GPT Image 2.5 上线 SeaArt,主打局部改图
演示用文字提示做定向修改,同时保持角色、场景和细节一致,走「生成—编辑—微调」的迭代流程。
GPT Image 2.5 just landed on SeaArt — and the editing workflow is getting seriously powerful. You can now make targeted changes with simple text prompts while keeping the character, scene, and important details consistent. ⚡ Faster generation = more iterations 🎯 Edit exactly what you want 🧠 Better consistency across edits 📸 More natural, high-fidelity results This is the kind of workflow that makes AI image creation feel less like generating from scratch and more like actually directing a creative project. Create → edit → refine → repeat. GPT Image 2.5 on SeaArt is worth testing. Check it out on @SeaArt_Ai https://x.com/SeaArt_Ai/status/2098323959129497818/video/1