Anime cosplayer phone screenshot: GPT Image 2 prompt

Generate an image styled like a vertical phone screenshot, with overall ratio close to 9:16. Slightly above center is a real-life cosplayer dressed as a 2D-anime character (character name). The figure is realistic but with slightly anime-touched features — fine skin, slightly larger eyes, gentle expression looking at the camera — sitting in a casual indoor scene like a café or bar counter, with props matching the scene in the background. At the very top, add a phone status bar UI with time, battery, signal, network icons, so the whole image looks like a phone screenshot. At the bottom, overlay a wide semi-transparent gal-game-style dialogue box. On the left of the box place an anime or chibi avatar matching the figure; on the right, lay out the text: line 1 in larger font shows the same character name; below it, 1–2 lines of warm healing-style Simplified Chinese dialogue suited to the character — auto-create the dialogue. Below the dialogue box, add an action bar imitating gal-game UI. Overall: high definition, rich detail, soft lighting, naturally fusing anime and live-action portrait.