Simple prompt trick gets chatbots to write generator‑ready image prompts, tests show

On June 2, 2026, Radhika Rajkumar highlighted a practical prompt technique from contributor Lance Whitney that gets chatbots to turn a short image idea into a generator‑ready prompt — a change that can yield richer inputs for image tools without requiring advanced prompt‑writing skill. The tactic matters for builders, designers and beginners because it can reduce iteration time and lower the barrier to producing detailed, usable prompts for different image generators.

Whitney’s example starts with a minimal user request: “I would like to create an image of a sunflower made of sheet metal in a pencil drawing style. Generate a prompt that I can use to request this image from [Nano Banana or ChatGPT Images].” The prescribed workflow is simple: supply only the core concept, ask the chatbot to expand it into a detailed prompt tailored to a target generator, and optionally request shorter or simpler variants if the result is too long.

The article reports tests with Gemini (used for Nano Banana). Gemini returned a focused, texture‑oriented prompt that emphasizes the metal construction and pencil shading: “A pencil drawing of a metal sunflower. The flower, its petals, the leaves, and the stem are composed of sculpted pieces of metal sheet. The texture of the metal is clear, with imperfections and visible welds. The drawing is detailed, with shading that creates depth and texture, emphasizing the three‑dimensional nature of the sculpture.” According to the report, feeding that output into Nano Banana produced a decidedly “metal‑ish” sunflower image.

ChatGPT produced a different, more granular prompt aimed at ChatGPT Images. The quoted output reads: “A detailed pencil drawing of a sunflower constructed from sheet metal. The petals are shaped from thin, slightly curved metal plates with visible seams, rivets, and subtle bends. The center of the sunflower is composed of layered metal textures resembling industrial components... Rendered in a realistic graphite pencil style with fine linework, cross‑hatching, and soft shading.” The article says this prompt yielded a successful image in ChatGPT Images.

The piece notes practical caveats: if a chatbot’s generated prompt is too long, you can ask for a shorter option; prompt behavior varies across tools; and companies frequently update or replace models, which changes outputs. Readers are pointed to the AI Model Release Tracker for monitoring model releases and differences between generations. For creators, the reported implication is straightforward: delegating prompt expansion to a chatbot can produce richer, generator‑ready text, reduce friction with safety filters or refused language, and save iteration time. Rajkumar calls the approach accessible to anyone and highlights particular utility for marketing and design work, while also urging creators to check workplace and copyright policies before using AI‑generated images.