Google — Structured prompts tuned for Gemini's multimodal and research strengths.
Use Gemini-optimized prompts for multimodal tasks (text plus image or video context), research that benefits from current information, and workflows tied into the Google ecosystem.
Strong at reasoning across text, image, and video context in the same prompt.
Good integration with search grounding for tasks that need current information.
Solid at structured data extraction and side-by-side comparison tasks.
Competitive long-context handling for large inputs.
Bad prompt:
“Compare these two products.”
Structured:
Role: You are a shopping assistant for a budget-conscious first-time buyer. Action: Compare Product A and Product B. Layout/Length: Output a table with columns: Feature, Product A, Product B, Winner. Follow with a 2-sentence recommendation.
A clean comparison table by feature, followed by a short, decisive recommendation — not a meandering paragraph comparison.
The generator doesn't have a dedicated Gemini mode yet — use Universal, which works well as a starting point.
Structured prompts tuned for GPT's conversational, general-purpose strengths.
Structured prompts tuned for Claude's careful reasoning and long-context strengths.