一只老虎一只老虎A minimalist prompt asking simply for "a tiger", generated as a still image in 1:1 format with GPT Image 2. Because the prompt carries almost no direction, the result relies entirely on the model's own interpretation.
Image prompts and examplesProduce a tiger image quickly with no constraints on style, setting, or composition. A prompt this short works well as a capability test, a generic visual to refine later, or a starting point for iterating toward a more specific look.
The generation uses a single phrase: "一只老虎", meaning "a tiger". No pose, background, lighting, or style is specified, so the model decides those elements freely. The 1:1 format produces a balanced square frame suited to social feeds or thumbnails. To reproduce the approach, enter a similarly brief description in GPT Image 2 while keeping the 1:1 ratio, then review the output. To push the result further, enrich the prompt step by step: first specify the tiger's attitude (walking, looking at the camera, resting), then the setting (jungle, snow, studio), and finally the visual treatment (realistic photography, illustration, watercolor).
With such a short prompt, pose, setting, lighting mood, and style are uncontrolled and can vary sharply from one render to the next. The 1:1 frame may crop wide environments such as a savanna or an extended forest. No text or brand elements can be added reliably without expanding the prompt.