AI Art Tools Face-Off

[Format] Horizontal, three-panel comic (three panels on the left → right). Anime-style + pop. Bright color scheme. The lines are clean. The background is a simple office.
[Characters] Emma (woman in her 20s, black bob hair, white shirt + navy jacket, expressive). All panels look the same and are unified in appearance.
[Text] Japanese is clearly displayed in an "easy-to-read" manner. No typos, garbled characters, or distortion. For text within speech bubbles, use the following phrases "word by word" as is.

First panel (left):
Emma explained with her finger.
Speech bubble letters (exact match):
"{argument name="First Frame Text" default="ChatGPT Images excels at illustrations, animations, and three-dimensional expressions"}"

Second panel (center):
Emma is handling the design work (layout and typography).
Speech bubble letters (exact match):
"{argument name="Second panel text" default="Nano Banana excels at Japanese text and design"}"

Panel 3 (right):
Comparison panel. The screen is split left and right, with the heading "ChatGPT Images" on the left and "Nano Banana" on the right.
Put a checklist on each item (in Japanese, short, easy to read).
Left (ChatGPT Images): 'Illustration', 'Anime', '3D Expression'
Right (Nano Banana): 'Japanese Text', 'Design', 'Layout'
Furthermore, a large "VS" is placed in the center to compare the two.

[NG] Unnecessary characters, arbitrary English additions, watermarks, signatures, illegible characters, and distorted Japanese.
[Image Quality] High resolution, sharp, suitable for social media posting.