Multimodal reasoning
Prompts that require the model to separate what it sees from what it infers.
24 curated prompts
Gemini Prompts
Gemini's strength is multimodal and large-context work. These prompts are written for tasks where the model must combine text, tables and media, then produce an auditable summary rather than a fluent guess.
Gemini promptsGoogle Gemini promptsmultimodal prompts
Spreadsheet and table work
Templates for reading tabular material and reporting findings with citations to rows.
Research summarisation
Prompts that keep claims traceable to the supplied sources.
Featured AI prompt templates
Copy-ready prompts selected from this topic cluster.
TextWriting
Lead Generator & Tracker (WordPilot.pro)
Prompt from prompts.chat: Lead Generator & Tracker (WordPilot.pro)
TextWriting
Deep Research Agent Role
Prompt from prompts.chat: Deep Research Agent Role
TextWriting
Accessibility Auditor Agent Role
Prompt from prompts.chat: Accessibility Auditor Agent Role
ImageImage Generation
Long Text to Xiaohongshu Multi-Image Handwritten Notes
Transform long-form articles into retro scholarly handjournal-style cards suitable for Xiaohongshu, combining academic rigor with artisanal aesthetics to generate multiple visual knowledge cards (450x600px) with single and batch download support.
TextWriting
Kintsugi Ethos Omni Ultimate Définitive Version Infinity Édition Balise PSY
Prompt from prompts.chat: Kintsugi Ethos Omni Ultimate Définitive Version Infinity Édition Balise PSY
TextWriting
Landing page copywriting
Prompt from prompts.chat: Landing page copywriting
TextWriting
technical-whitepaper-writer
Prompt from prompts.chat: technical-whitepaper-writer
TextWriting
Photography Trip Planning — Research-Backed Itinerary Builder
Prompt from prompts.chat: Photography Trip Planning — Research-Backed Itinerary Builder
TextWriting
AI Provider Research Expert
Prompt from prompts.chat: AI Provider Research Expert
TextWriting
Unified Research and Source Analysis Prompt
Prompt from prompts.chat: Unified Research and Source Analysis Prompt
Do multimodal prompts need special structure?
Yes. Separate the observation instructions from the reasoning instructions, and require the model to flag anything it cannot actually see.
How do I stop a model from inventing visual details?
Ask for a description of what is present first, then a separate inference section, and require explicit uncertainty markers.
Can these prompts be used for document review?
Yes. They are commonly used for extracting structured findings from reports, forms and slides.