Gemini AI Prompts: Examples, Templates & How to Write Better Ones in 2026
40 min read
Google changed the Gemini lineup four times over the summer, and the prompt style that works changed with it.
A strong Gemini AI prompt names the task, gives the relevant context, sets the output format, and adds one clear limit. Paste this one and you will see the difference: "Summarize this report for a non-technical manager in five bullets, highlight the main risks, and keep it under 150 words."
That same structure carries across the whole lineup, from Gemini 3.8 Flash for fast multi-step work to Nano Banana 2 for images and Lyria 3.5 for music.
Below you get 100+ prompts you can copy and paste by task, the reusable template behind all of them, the model guide for September 2026, and a free generator that builds a prompt around your own work.
Turn Your Idea Into a Copy-Ready Gemini Prompt 👇
Describe the task in your own words. The generator returns a structured prompt you can paste straight into Gemini.
Best Gemini AI Prompts to Copy and Paste
These ten prompts cover the tasks people bring to Gemini most often. Copy any one of them, swap the text inside the square brackets for your own details, and paste it into Gemini. Each card names the job it does, the part you should change, and the model that handles it best.
💡 The bracket rule. Everything inside [square brackets] is a blank you fill in. Everything outside the brackets is the instruction Gemini follows, so leave it alone the first time you run a prompt.
✍️ Writing: blog section drafting
Best for: turning an outline point into publishable copy.
Write a 300-word section for a blog post about [topic]. Audience: [describe your reader]. Open with the direct answer in the first sentence. Use short paragraphs and one bulleted list. Avoid filler phrases and marketing language.
What to customize: the topic, the reader, and the word count.
Recommended model: Gemini 3.1 Pro for research-heavy topics, Gemini 3.8 Flash for speed.
📄 Summarize: long document to five bullets
Best for: reports, contracts, and meeting transcripts you have no time to read.
Summarize the attached [document type] into five bullets for a [job title]. Cover the decisions made, the money involved, and anything that needs my approval. Keep the whole summary under 150 words. Flag anything ambiguous instead of guessing.
What to customize: the document type and the job title of the person reading it.
Recommended model: Gemini 3.1 Pro, because long uploads reward careful reading.
🔍 Research: source-checked briefing
Best for: getting up to speed on a topic before a meeting.
Research [topic] and give me a briefing with four sections: what is settled, what is disputed, what changed in the last six months, and what I should ask about. Cite a source for every claim. Separate verified facts from your own inference.
What to customize: the topic and the four section names.
Recommended model: Gemini 3.1 Pro with Deep Research turned on.
📧 Email: client update that lands
Best for: the weekly status email you keep putting off.
Write a client update email for [project name]. Status: [where things stand]. Next step: [what happens next]. Blocker: [what you need from them]. Warm but brief tone, under 150 words, subject line included. End with one yes-or-no question.
What to customize: the three bracketed lines. Fill them with rough notes, not polished sentences.
Recommended model: Gemini 3.8 Flash.
🎓 Studying: revision plan from your own notes
Best for: exam week, when you have material but no plan.
Here are my notes for [subject]: [paste notes]. Identify the five highest-priority topics. Then build a 7-day revision schedule with daily study blocks, three active-recall questions per block, and one short quiz per day. Output as a table.
What to customize: the subject and the number of days you actually have.
Recommended model: Gemini 3.1 Pro.
📊 Data analysis: trends and outliers
Best for: a spreadsheet you need to explain to someone else.
Analyze this dataset: [paste or upload]. Identify the three clearest trends, any figure that looks wrong, and what the numbers suggest for [the decision you face]. Give the answer first, then the working. Say so plainly if the data cannot support a conclusion.
What to customize: the decision at the end. That line does most of the work.
Recommended model: Gemini 3.8 Flash for speed, Gemini 3.1 Pro for large files.
💻 Coding: multi-step build
Best for: work that needs planning before any code gets written.
Step 1: read the code I am pasting and describe what it does.
Step 2: list the bugs and the risky assumptions.
Step 3: rewrite it with the fixes applied.
Step 4: explain each change in one line.
Finish each step before starting the next and label every step.
What to customize: the four steps. Add or remove steps to match the job.
Recommended model: Gemini 3.8 Flash, which Google built for long-horizon coding
🖼️ Image: Gemini AI photo prompt
Best for: a specific picture rather than a lucky one.
Create a [shot type] of [subject]. Setting: [where]. Lighting: [natural window light / studio softbox / golden hour]. Composition: [close-up / wide / overhead]. Style: [editorial photography / 2D illustration / product render]. Keep skin tones natural and the text legible.
What to customize: all five labels. Naming lighting and composition changes the result more than any adjective.
Recommended model: Nano Banana 2 for speed, Nano Banana Pro for final assets.
🎬 Video: short clip brief
Best for: social clips, product shots, and explainer scenes.
Create an 8-second video of [scene]. Motion: [slow push-in / fast cuts / static]. Camera angle: [low 45-degree / handheld close-up / wide]. Duration: 8 seconds. Style: [cinematic commercial / clean 2D animation / documentary]. Mood: [calm / energetic / dramatic].
What to customize: the scene and the mood. Keep the duration short.
Recommended model: Veo 3.1 through the Gemini API and Google AI Studio.
🗓️ Planning: 90-day plan with owners
Best for: launches, migrations, and anything with a deadline.
Build a 90-day plan for [project]. Break it into three 30-day phases with weekly milestones. For each milestone give the owner role, one risk, and one success measure. Format as a table. Flag any milestone that depends on something outside my control.
What to customize: the project and the phase length.
Recommended model: Gemini 3.1 Pro.
📌 Want the rest? The full 100+ library further down covers images, writing, research, work, students, video and music, grouped by what you want to produce.
Gemini rewards structure more than phrasing, so the same request written two ways can return very different answers. Prompts written for other chatbots often underperform here for that reason.
Gemini 3 processes text, images, audio and video natively, and it supports up to 1 million tokens of context. That means it can work through hundreds of pages of documents, whole codebases, or long video transcripts in one pass.
As of September 2026, Gemini 3.8 Flash is the newest model in the Flash line. Google describes it as its most intelligent workhorse model and says it was built for long-horizon software engineering, autonomous agents and multi-step reasoning. Those are Google's own claims about its own model.
Three habits change your results more than anything else:
Give context before instructions. When you upload a document or paste background, Gemini weights that material heavily. Clean, relevant, well-ordered source text produces better output than a longer command.
Set limits as numbers, not adjectives. "Keep it short" leaves room for interpretation. "Maximum 150 words" does not. The same applies to format, exclusions and sequence.
Ask for depth when you want depth. Gemini defaults to a compact answer. If you want a step-by-step breakdown, say so in the prompt, or you will get a summary.
How to Write a Good Prompt for Gemini
A good prompt for Gemini gives five things: the task, the context, the audience, the output format, and the limits. Add one worked example when tone matters, and split anything complicated into numbered steps. Most weak answers come from a missing audience or a missing format, not from weak wording.
Here is the same request, before and after.
❌ Weak: "Write something about our new pricing."
✅ Strong: "Write a 120-word announcement about our new pricing for existing customers who are on the old plan. Explain what changes, what stays the same, and what they need to do. Plain language, no marketing adjectives, one bulleted list."
The strong version took twenty extra seconds to type. It removes every decision Gemini would otherwise make for you.
Google's own prompt engineering guide covers the same ground for developers, and our guide on how to write good AI prompts walks through the four-part build with before-and-after examples.
💡 The step rule. When one prompt carries two unrelated jobs, output quality drops on both. Write "Step 1... Step 2..." and tell Gemini to finish each step before starting the next.
The 3-Part Gemini Prompt Formula (Intent + Context + Constraints)
The most common question beginners have is: what prompt should I use for Gemini?
The answer depends on your output and this three-part formula makes that decision simple every time.
Most people type into Gemini the same way they'd search Google. That works for simple lookups, but it breaks down fast when you need something genuinely useful.
Gemini responds to structure, and once you give it that structure, the quality of your output changes immediately.
The formula has three parts. Every strong Gemini prompt uses all three.
Part | What it means | Example |
|---|---|---|
Intent | The exact output you want | "Write a 5-bullet summary" |
Context | Your role, situation, audience | "I'm a PM presenting to a non-technical board" |
Constraints | Format, length, tone, exclusions | "Max 120 words, no jargon" |
Here's what that looks like in practice:
❌ Weak prompt: "Summarize this document."
✅ Strong prompt: "I'm a product manager presenting to a non-technical board. Summarize this 40-page product roadmap into 5 bullet points. Focus on business impact and timelines. Avoid technical terminology. Max 120 words."
Same task. Completely different output quality.
The strong version tells Gemini the intent (5-bullet summary), the context (non-technical board presentation), and the constraints (no jargon, 120-word cap).
This is the foundation of every good Google Gemini prompt template and once you internalize it, writing prompts gets much faster.
Reusable Gemini Prompt Template
Once you understand the 3-part formula, this template makes it easy to apply every time.
Below is a Gemini prompt copy paste template you can reuse for any task. Fill in the brackets and you're done.
📋 The reusable Gemini prompt template
Best for: any task, on any Gemini model.
[What you want Gemini to produce]. I am [your role/situation] and this is for [audience/purpose]. Output should be [bullet points/table/paragraph/numbered list]. Keep it [length + tone]. Do not include [specific exclusions].
What to customize: all five brackets. The last one matters most, because exclusions are what most people forget.
A filled-in example looks like this:
"Write a project status update. I am a marketing manager and this is for a weekly client report. Output should be bullet points. Keep it under 100 words and professional in tone. Do not include internal team names or budget figures."
That's it. Same structure, any task.
If you're on Gemini 3.1 Pro, you can also select "Pro" or "Thinking mode". This activates deeper reasoning for complex research or multi-step tasks and it makes a noticeable difference on detailed outputs.
Fill the template yourself, or describe your task once and let the generator build the prompt around it.
Need a prompt built around your specific task instead of a blank template? Phrasly's free AI Prompt Generator fills in all the details for you automatically. Try it for free 👇
Current Gemini Models and How to Prompt Them (September 2026)
Gemini 3.8 Flash is the newest model in the lineup, released on September 2, 2026, and Gemini 3.1 Pro remains the reasoning flagship. Gemini 3.5 Pro was announced for June and never shipped, so 3.1 Pro is still the model to reach for on hard reasoning work.
A single prompt does not perform equally well across all of them. Each model rewards a different shape of instruction, and the table below maps the shape to the model.
Model | Released | Available through | Best for | Prompt style |
|---|---|---|---|---|
Gemini 3.8 Flash | Sep 2, 2026 | Gemini app (AI Pro and Ultra), AI Mode in Search, Gemini in Google Sheets, Gemini API, AI Studio, Antigravity | Long-horizon coding, autonomous agents, multi-step reasoning | Numbered task chain with labeled outputs |
Gemini 3.6 Flash | Jul 21, 2026 | Gemini app, Gemini API, AI Studio | Everyday work at lower cost than 3.5 Flash | Plain task plus format, no step chain needed |
Gemini 3.1 Pro | Feb 2026 | Gemini app, Gemini API, AI Studio | Deep reasoning, long uploads, careful analysis | Source material first, task instruction last |
Gemini 3.5 Flash-Lite | Jul 21, 2026 | Gemini API, AI Studio | High-volume tagging, classification, extraction | Short prompts with fixed labels |
Nano Banana 2 (Gemini 3.1 Flash Image) | GA May 28, 2026 | Gemini app, Google Search and Lens, Gemini API, AI Studio | Fast image creation and everyday edits | Subject, style, lighting, composition |
Nano Banana Pro (Gemini 3 Pro Image) | GA May 28, 2026 | Gemini app regenerate on paid plans, Gemini API | Studio-quality visuals and legible text in images | Detailed scene with composition and specs |
Veo 3.1 / Veo 3.1 Lite | Lite preview Mar 31, 2026 | Gemini API, Google AI Studio, Google AI Pro and Ultra plans | Short video clips and rapid iteration | Scene, motion, camera angle, duration, style |
Lyria 3.5 | Sep 3, 2026 (API preview), Sep 4 (Gemini app) | Gemini app, Gemini API, AI Studio, Google Flow Music, Google Vids | Full songs and short music clips | Genre, tempo, mood, instrumentation, duration |
Gemma 4 | Apr 2, 2026 | Google AI Studio, Gemini API, open-weight download | Self-managed and on-device deployments | Standard prompts tuned to your own setup |
Model names, release dates and availability reflect Google's own live pages and the Gemini API changelog as checked in September 2026. Google is shipping new Gemini models every few weeks, so confirm the current lineup on Google's model pages before relying on any row above.
⚠️ Availability is not the same everywhere. The same model name behaves differently depending on your account, plan, country and interface. The Gemini app, the Gemini API, Google AI Studio, Google Workspace and Vertex AI each expose a different set. Check what your own account actually shows before you plan a workflow around a model.
The release-history narration that used to run here has been cut. It carried eight model launches that no longer describe the current lineup, and the table above replaces all of it.
Deep Research and Analysis Prompts
These prompts handle the work that used to take hours: large documents, competitive maps, and decisions with several moving parts. Run them on Gemini 3.1 Pro.
Long document analysis
Best for: finding every mention of one theme across a long report.
I'm uploading a 150-page industry report. Find every mention of [topic]. For each one, tell me where it appears, why it matters in context, and give me a timeline of key developments. Output as a structured list.
What to customize: the topic and the page count.
Competitor analysis
Best for: a side-by-side view you can take into a strategy meeting.
Act as a market strategist. Analyze the positioning, pricing model, and key weaknesses of [Company A, B, C] in [industry]. Output as a side-by-side comparison table with one row of strategic recommendations at the bottom.
What to customize: the company names and the industry.
Complex decision-making
Best for: choices with more than two variables.
Help me evaluate these 3 business options using a decision matrix. Criteria: cost, impact, risk, and 12-month scalability. Think through each criterion before scoring. Present the final matrix as a table with a summary recommendation.
What to customize: the criteria. Swap in whatever your decision actually turns on.
Each of these follows the Intent + Context + Constraints formula. The explicit output format instruction at the end is what separates a clean, usable result from a wall of text.
Gemini 3.1 Pro Prompts for Agentic and Long-Context Tasks
Gemini 3.1 Pro handles multi-step workflows, long uploaded documents, and multimodal inputs. Place any examples or reference material before your task instruction, because the model prioritizes earlier context.
Multi-step project planning
Best for: turning an uploaded brief into a dated plan.
You are managing a product launch. Using the document I've uploaded as your source, build a 90-day plan broken into weekly milestones. For each milestone, include: owner role, risk flags, and success metric. Format as a structured table.
What to customize: the project type and the plan length.
Video and visual analysis
Best for: briefing a shoot from a reference clip.
Analyze this video clip. Identify the visual motifs, dominant color palette, and pacing style. Then suggest 3 real-world locations that match this aesthetic for a brand shoot.
What to customize: the number of locations and the output you need.
Agentic workflow prompt
Best for: a sequence of related pieces written in one pass.
I need to build an onboarding email sequence for a SaaS product. Step 1: write a welcome email. Step 2: write a Day 3 check-in. Step 3: write a Day 7 feature highlight. Use a friendly but professional tone. Max 150 words per email.
What to customize: the product and the number of steps.
Gemini 3.8 Flash Prompts for Agentic Loops
Gemini 3.8 Flash runs a numbered chain in one pass, so a four-part job does not need four separate prompts. Google says the model executes extra reasoning steps and calls tools repeatedly on complex tasks, which is why labeling each step matters.
Agentic loop: competitive analysis
Best for: research, extraction, mapping and writing in a single run.
You are running a competitive analysis loop. Step 1: research [Company A]. Step 2: extract their pricing model. Step 3: map it to our positioning. Step 4: write a 100-word summary. Execute each step before moving to the next, and output each step's result labeled clearly.
What to customize: the four steps and the company.
Recommended model: Gemini 3.8 Flash.
📌 Cost note from Google. Google tells developers that 3.8 Flash may use more tokens at higher effort levels, and suggests lower effort settings or staying on 3.7 Flash when compute efficiency matters more than depth. That is Google's own guidance, published with the model
100+ Best Prompts for Gemini AI (By Output Type)
This is the full 100+ library, grouped by what you want to produce rather than by who you are. Jump to images, writing, research, work, students, video or music, copy the prompt, swap the [brackets], and you are done.
Most people don't sit down thinking, "I'm a content marketer, what should I prompt?" They think, "I need an email written" or "I need this report summarized in two minutes." That's exactly how this section is built.
Every prompt below is formatted for direct copy-paste, just swap the [brackets] for your specifics.
Photo Prompts for Gemini AI: Nano Banana 2 Edition
Image generation runs on Nano Banana 2 by default across the Gemini app, Google Search AI Mode and Google Lens. Nano Banana 2 first launched as Gemini 3.1 Flash Image Preview in February 2026, and both Nano Banana 2 and Nano Banana Pro reached general availability in May 2026.
Gemini now has two image lanes. Nano Banana 2 is the faster choice for quick creation, edits and everyday visual prompts. Nano Banana Pro is the one for polished, detailed, higher-control output, and paid subscribers can regenerate a result with it.
Because the model is built into Gemini's multimodal architecture, it can pull from real-world knowledge to generate more accurate scenes, diagrams and visuals. You can describe the composition, reference visual context, and ask for precise edits rather than a vague "make this better".
💡 Name five things in every image prompt: subject, setting, lighting, composition, and style. Prompts that name all five beat long adjective lists almost every time
Gemini Cinematic & Portrait Prompts
Moody noir
Best for: dramatic, high-contrast character shots.
Create a cinematic noir portrait of a detective standing under a streetlight at night. High contrast lighting, deep shadows, rain-soaked pavement reflections, 50mm lens style, dramatic film lighting.
What to customize: the character and the lens style.
Golden hour backlit
Best for: warm outdoor portraits.
Generate a portrait of a traveler on a cliff during golden hour. Warm backlighting, soft lens flare, shallow depth of field, natural skin tones, DSLR photography style.
What to customize: the subject and the location.
8K studio portrait
Best for: clean, detailed headshots on a plain background.
Create an ultra-detailed studio portrait of a professional speaker against a neutral background. Softbox lighting, crisp facial details, high dynamic range, 8K resolution.
What to customize: the profession and the background color.
Vintage film grain
Best for: an analog look without a filter.
Generate a portrait with vintage film grain texture, muted color palette, analog camera feel, and subtle imperfections similar to 1970s film photography.
What to customize: the decade and the color palette.
Cyberpunk neon
Best for: stylized night scenes.
Create a cyberpunk city portrait with neon lights reflecting on wet streets, futuristic clothing, holographic signage, and cinematic color grading.
What to customize: the city details and the color grading.
Gemini Professional & LinkedIn Headshot Prompts
If you need something polished and professional, these work well.
Corporate headshot
Best for: a LinkedIn profile photo.
Generate a professional LinkedIn headshot of a business consultant in formal attire. Neutral studio background, soft lighting, natural skin tones, sharp focus.
What to customize: the job title and the attire.
Creative industry headshot
Best for: designers, writers and founders.
Create a modern headshot for a creative professional. Bright studio lighting, minimalist background, casual professional clothing, editorial photography style.
What to customize: the clothing and the lighting.
Outdoor natural light
Best for: a warmer, less formal profile photo.
Create a professional outdoor headshot in natural daylight. Soft background blur, warm natural tones, relaxed but professional expression, high resolution.
What to customize: the setting and the expression.
Professional portrait prompt for men
Best for: professional headshot-style images for men.
Generate a professional portrait of a male business professional for LinkedIn. Outfit: tailored navy blazer, white shirt, no tie. Background: clean modern office with soft blur. Lighting: natural window light, realistic skin texture, confident but approachable expression, sharp focus, high-resolution editorial style.
What to customize: the outfit and the background.
Image Quality Enhancement Prompt
Image quality enhancement
Best for: when you already have a photo and want it improved without changing the picture.
Enhance this image while keeping the original subject, pose, and composition unchanged. Improve sharpness, lighting balance, skin texture, color accuracy, and background clarity. Remove minor noise or blur, but do not make the image look artificial or over-edited. Keep it realistic and high-resolution.
What to customize: nothing. Upload the image and run it as written.
Background Replacement & Scene Prompts
Already have a subject photo? These prompts let you swap the environment while keeping the subject intact.
Misty forest
Best for: a soft, atmospheric backdrop.
Keep the subject unchanged but replace the background with a misty pine forest at sunrise. Soft atmospheric lighting, cinematic mood.
What to customize: the location and the time of day.
Urban rooftop
Best for: a city skyline backdrop.
Keep the subject centered and place them on a modern city rooftop at sunset with skyline and warm cinematic lighting.
What to customize: the city and the light.
Minimalist studio
Best for: product shots and clean profile photos.
Replace the background with a clean white studio backdrop, soft shadow lighting, editorial photography style.
What to customize: the backdrop color.
Infographic & Data Visualization Prompts
Because Nano Banana 2 can render text within images, it's also surprisingly good at diagrams and visual content.
Process flowchart
Best for: explaining a sequence in one image.
Create a clean infographic showing a 5-step marketing funnel. Flat design, labeled steps, connecting arrows, modern color palette.
What to customize: the number of steps and the labels.
Flat-lay diagram
Best for: workflow slides and internal decks.
Generate a flat-lay product development workflow diagram. Minimalist icons, labeled stages, white background, modern corporate design.
What to customize: the stages.
Dashboard mockup
Best for: showing a product idea before it is built.
Design a digital marketing dashboard mockup with analytics charts, labeled metrics, and a clean modern UI style.
What to customize: the metrics on screen.
Gemini Prompts for Writing & Content
Writing takes up most of the working day for most people: emails, reports, briefs, blog posts, client updates. These prompts cut the time without cutting the quality.
Gemini also embeds an invisible SynthID watermark in its text output, and our guide on how to remove the Gemini text watermark explains what that watermark is and how it affects your drafts.
Gemini Blog & Article Writing Prompts
SEO article
Best for: a full draft you will edit rather than publish raw.
Write a 1,500-word SEO article targeting the keyword [keyword]. Audience: [describe your reader]. Structure it with an intro, 4 H2 sections, and a conclusion. Each section should be 250-300 words. Use a conversational but authoritative tone. Avoid filler phrases.
What to customize: the keyword, the reader, and the word count.
Thought leadership LinkedIn article
Best for: posts written to start a conversation.
Write a 700-word LinkedIn article from the perspective of a [your role] sharing an unpopular opinion about [industry topic]. Open with a bold statement, support it with 3 arguments, and end with a question to drive comments. Professional but direct tone.
What to customize: the role and the opinion.
Blog outline with stat placeholders
Best for: planning before you write.
Create a detailed blog outline for an article titled '[your title]'. Include an H1, 5 H2s with 2-3 H3s each, and note where statistics or data points should be added. Target audience: [describe reader]. Focus keyword: [keyword].
What to customize: the title and the heading counts.
Gemini Professional Bio & LinkedIn Prompts
150-word LinkedIn bio
Best for: a LinkedIn About section or a company website page.
Write a 150-word professional bio for [name], a [job title] at [company]. Highlight [2-3 key achievements or specialties]. Tone should be confident but approachable. Written in third person. Suitable for a LinkedIn About section or company website.
What to customize: the achievements. Those carry the bio.
LinkedIn headline variations
Best for: testing five angles before you commit to one.
Write 5 LinkedIn headline variations for a [job title] who specializes in [niche]. Each should be under 220 characters. Mix formats, one achievement-based, one keyword-heavy, one value proposition, one personality-led, one role + result.
What to customize: the job title and the niche.
Gemini Email & Communication Prompts
Client progress update
Best for: the recurring status email.
Write a professional client update email for [project name]. Current status: [brief update]. Next steps: [what's happening next]. Tone: warm but concise. No more than 150 words. Subject line included.
What to customize: the status and next steps.
Polite follow-up on unanswered proposal
Best for: chasing without sounding like chasing.
Write a polite follow-up email for a proposal sent [X days] ago to [client type]. Keep it short, 3 sentences max. Don't sound pushy. End with a low-friction call to action like a yes/no question.
What to customize: the number of days and the client type.
Internal team update (Slack-style)
Best for: stand-up notes and channel updates.
Write a Slack-style internal update for the [team name] team about [project/topic]. Keep it under 80 words. Use short sentences, no corporate jargon. Include: what's done, what's in progress, and any blockers. Casual but professional tone.
What to customize: the team and the topic.
If your Gemini output still sounds machine-written, our guide on how to humanize AI text covers what to change to make the final draft read naturally.
Gemini Prompts for Research & Summarization
Long-context handling is where Gemini earns its place. It works through dense, lengthy documents better than almost any AI tool right now, and these prompts make that practical for everyday work.
Gemini Document & Report Summarization Prompts
Executive summary
Best for: briefing leadership in under a page.
Summarize this [document/report] into a 150-word executive summary for senior leadership. Focus on key decisions, business impact, and recommended next steps. Avoid technical language. Use clear, direct sentences.
What to customize: the document type and the audience.
Key insight extraction
Best for: pulling the five things that matter out of fifty pages.
Read this report and extract the 5 most important insights. For each one, explain what it means in plain language and why it matters to [your industry/role]. Output as a numbered list.
What to customize: the industry or role.
Gemini Academic & Literature Research Prompts
Literature review structure
Best for: planning a review before you read everything.
Help me structure a literature review on [topic]. Identify the major themes, key debates, and gaps in existing research. Output as an outline with subheadings I can write to. Academic tone, suitable for a [undergraduate/postgraduate] level paper.
What to customize: the topic and the academic level.
Research question formulation
Best for: narrowing a topic that is still too broad.
I'm researching [broad topic]. Help me narrow this into 3 focused research questions that are specific, measurable, and suitable for a [dissertation/journal article/report]. Explain why each question is worth pursuing.
What to customize: the broad topic and the output type.
Citation integration and evidence mapping
Best for: matching evidence to the argument it supports.
I have the following evidence points: [paste your notes]. Help me map them to the argument I'm building about [topic]. Suggest where each piece of evidence fits strongest and flag any gaps I still need to fill.
What to customize: your notes and the argument.
Literature synthesis
Best for: turning several uploaded sources into one structured report.
I am a [researcher/student] studying [topic]. Analyze the following sources I've uploaded and produce a structured literature synthesis covering key arguments, areas of consensus, and unresolved debates. Output as a formatted report with sections and bullet points.
What to customize: your role and the topic.
Gemini Notes & Action Items Prompts
Raw notes to structured summary
Best for: meeting notes you typed too fast to read back.
Here are my raw meeting notes: [paste notes]. Convert them into a structured summary with three sections: Key Decisions, Action Items, and Open Questions. Assign an owner and deadline to each action item where mentioned. Keep the total under 200 words.
What to customize: the three section names.
Gemini Prompts for Work & Productivity
These handle the strategic and operational tasks that take the most time and attention.
Gemini Strategy & Planning Prompts
SWOT analysis
Best for: a first pass on a product or market position.
Run a SWOT analysis for [company/product/idea]. Research the current market context where relevant. Output as a clean 2x2 table with 3-4 bullet points in each quadrant. Keep each point specific and actionable, not generic.
What to customize: the company or product.
90-day launch plan
Best for: anything with a deadline and more than one owner.
Create a 90-day launch plan for [product/project]. Break it into three 30-day phases with weekly milestones, key risks per phase, and one success metric per milestone. Format as a structured table.
What to customize: the project and the phase length.
Customer personas
Best for: messaging work that needs a reader in mind.
Create 3 customer personas for [product/service]. For each persona include: demographics, primary pain points, buying motivations, and 2 messaging angles we could use to reach them. Output each persona as a separate labeled section.
What to customize: the product and the number of personas.
Gemini Data & Reporting Prompts
Data analysis and trend extraction
Best for: a dataset you need to explain, not just read.
Analyze this dataset: [paste data or upload file]. Identify the top 3 trends, any anomalies worth flagging, and what the data suggests for [business decision]. Output as a short report with a summary at the top.
What to customize: the business decision at the end.
Report-to-presentation conversion
Best for: turning a document into a deck outline.
Convert this report into a 10-slide presentation structure. For each slide include: a title, 3 bullet points max, and a suggested visual or chart type. Audience: [describe]. Tone: [professional/executive/client-facing].
What to customize: the slide count and the audience.
Gemini Google Workspace-Specific Prompts
These prompts only work inside Google Workspace, where Gemini has direct access to your live documents and email threads.
Email thread follow-up
Best for: picking up a conversation you have lost track of.
Look at my last 5 emails to [client name]. Draft a follow-up email summarizing our open action items and next steps. Professional tone, under 150 words.
What to customize: the client name and the email count.
Google Doc to task list
Best for: converting a decision document into work.
Summarize the key decisions from this Google Doc and create a task list with owners, deadlines, and priority levels. Format as a table.
What to customize: the columns you want in the table.
Best Instructions for Gemini: Reusable Templates
Permanent instructions for Gemini live in Personal Context, under a setting called Instructions for Gemini, inside the Gemini app's Personal Intelligence settings. You set your default response preferences once, and Gemini applies them to every future chat. You can switch them on or off from the Gemini app or the web app whenever you want.
Use one when you want Gemini to write, research, summarize or format the same way every time. Here are three you can paste straight into the Instructions field.
1. Professional writer persona
Best for: emails, blogs, reports, rewrites, and professional communication.
Act as a clear, practical professional writer. Write in short paragraphs, avoid generic AI phrases, and lead with the most useful answer first. Keep the tone confident, simple, and human. When editing text, improve clarity without changing the original meaning. Avoid overexplaining unless I ask for detail.
What to customize: the tone words at the end.
2. Research assistant
Best for: research summaries, academic work, competitor research, and source-based writing.
Act as a careful research assistant. When I ask about a topic, separate verified facts from assumptions. Prefer official sources, academic sources, and primary documentation where possible. Summarize findings in a structured format with headings, bullets, and clear next steps. If information is uncertain or outdated, say so directly.
What to customize: the source types you trust.
3. Meeting summarizer
Best for: Google Docs notes, Gmail threads, project updates, and team meetings.
Act as my meeting summarizer. Turn messy notes, transcripts, or email threads into a clean summary with four sections: Key Decisions, Action Items, Owners, and Open Questions. Keep action items specific and deadline-focused. Use a concise, professional tone.
What to customize: the four section names.
Set these once in Gemini's Instructions field, and every response will follow your defaults.
Gemini Prompts for Students & Academic Use
These Gemini prompts for students turn Gemini into a study partner. Use them to structure essays, prepare for exams, simplify difficult concepts, map research evidence, and organize ideas before you start writing.
Essay Structuring
1. Compare/contrast outline
Best for: when you know your topic but need a clean essay structure.
Create a compare/contrast essay outline on [Topic A] vs [Topic B]. Include an intro with a clear thesis, 3 body sections each covering a distinct point of comparison, and a conclusion. Academic tone, suitable for a [high school/university] level assignment.
What to customize: the two topics and the level.
2. Argumentative essay thesis and structure
Best for: persuasive essays, debate assignments, and papers where your argument needs to be clear.
Help me build an argumentative essay on [topic]. Start by suggesting 3 possible thesis statements: one moderate, one bold, and one nuanced. Then outline the strongest one with 4 supporting arguments and a counterargument I should address. Target length: [word count].
What to customize: the topic and the target length.
Study & Exam Prep
3. Exam revision plan
Best for: when finals are close and you need a realistic plan from your own class material.
Create a study plan for my [exam name/subject] using the notes I provide. First, identify the highest-priority topics. Then create a 7-day revision schedule with daily study blocks, active recall questions, flashcard topics, and one short practice quiz per day. Output as a table.
What to customize: the subject and the number of days.
4. Concept-to-analogy for difficult subjects
Best for: biology, chemistry, economics, statistics, coding, or any subject that feels abstract.
Explain [difficult concept] using a simple real-world analogy. Then explain where the analogy works and where it breaks down. After that, give me 3 practice questions to check whether I actually understand it.
What to customize: the concept.
Research & Applications
5. Research citation mapping
Best for: literature reviews, research essays, and assignments where your evidence needs to match your claims.
I am writing a research paper on [topic]. Here are my sources and notes: [paste notes]. Map each source to the part of my argument where it fits best. For each source, explain what claim it supports, what evidence it provides, and whether I need more recent or stronger support.
What to customize: the topic and your notes.
6. College application essay outline
Best for: when you want structure and brainstorming help while keeping the final essay in your own voice.
Help me outline a college application essay about [personal experience/theme]. Do not write the essay for me. Instead, create 3 possible essay angles, a clear opening hook for each, the main lesson or growth arc, and a closing idea that feels personal rather than generic.
What to customize: the experience or theme.
For more study-specific prompts, see our ChatGPT Prompts for Students guide. The infographic and diagram prompts from the photo section above also work well for presentations, reports and academic visuals.
New Trending Gemini Prompts: Updated September 2026
Three things changed in the Gemini lineup since this section was last written. Gemini 3.8 Flash arrived on September 2, 2026 for fast agentic work, Lyria 3.5 reached the Gemini app on September 4, 2026 for music, and video generation still runs on Veo 3.1. The prompts below are built for those three.
Best Gemini 3.8 Flash Prompts for Agentic Tasks
Gemini 3.8 Flash suits multi-step work where you want one pass instead of five follow-up prompts. Number the steps and ask for each result to be labeled.
1. Content refresh audit
Best for: deciding what to update on a page that has gone stale.
Review this blog post: [paste or upload content]. Step 1: Identify outdated sections. Step 2: List missing 2026 information. Step 3: Suggest SEO updates for headings and keywords. Step 4: Rewrite the intro in a clearer, more current style. Output as a table plus final revised intro.
What to customize: the year and the four steps.
2. Workflow automation
Best for: documenting a process you repeat every week.
Create a repeatable workflow for [task]. Break it into trigger, input, process, quality check, and final output. Include tools needed, possible failure points, and a 5-step checklist I can reuse.
What to customize: the task and the checklist length.
3. Research brief
Best for: a briefing document with a short summary at the end.
Build a research brief on [topic]. First, summarize the current state of the market, then identify key trends, risks, expert viewpoints, and unanswered questions. Output in sections with bullets. Keep the final summary under 150 words.
What to customize: the topic and the section names.
4. Decision matrix
Best for: choosing between three options you keep going back and forth on.
Help me choose between [Option A], [Option B], and [Option C]. Score each option from 1 to 5 on cost, speed, risk, effort, and long-term value. Explain each score briefly, then recommend the best option.
What to customize: the options and the scoring criteria.
Gemini Prompts for Veo Video Generation
Video generation runs on Veo 3.1, with Veo 3.1 Lite as the cost-efficient preview model for high-volume work. You reach both through the Gemini API and Google AI Studio rather than the standard chat box, and Veo 3.1 is also included with the Google AI Pro and Ultra plans. Clips run short, so write each prompt with a few seconds in mind.
Every Veo prompt works best when it names five things: scene, motion direction, camera angle, duration, and style.
1. Product video
Best for: a premium product shot for a landing page.
Create an 8-second cinematic product video for [product]. Scene: product on a premium studio surface. Motion: slow push-in. Camera angle: low 45-degree angle. Duration: 8 seconds. Style: realistic commercial lighting, soft shadows, premium brand feel.
What to customize: the product and the surface.
2. Social short
Best for: vertical clips for Reels, Shorts and TikTok.
Generate a 7-second vertical video for [topic]. Scene: fast opening hook, product or idea shown in action, quick final payoff. Motion: fast cuts and smooth transitions. Camera angle: handheld close-ups. Style: energetic, modern, mobile-first.
What to customize: the topic and the pacing.
Gemini Prompts for Lyria Music Generation
Lyria 3.5 is Google's current music generation model. It entered public preview in the Gemini API on September 3, 2026 and reached the Gemini app a day later, alongside Google Flow Music, Google Vids and Google AI Studio. It builds complete tracks with verses, choruses and bridges, and every track carries a SynthID watermark.
A strong Lyria prompt names five things: genre, tempo, mood, instrumentation, and length.
1. Product jingle
Best for: short brand audio and ad beds.
Create a 30-second upbeat product launch jingle for [product]. Genre: bright pop-commercial. Tempo: 128 BPM. Mood: optimistic and energetic. Instrumentation: drums, synth bass, hand claps, clean guitar, and short vocal hook. Duration: 30 seconds.
What to customize: the product, the tempo and the instruments.
2. Study music
Best for: background loops with no vocal distraction.
Generate a 30-second lo-fi study loop. Genre: lo-fi hip-hop. Tempo: 78 BPM. Mood: calm, focused, and warm. Instrumentation: mellow piano, soft vinyl crackle, brushed drums, warm bass, and ambient texture. Duration: 30 seconds.
What to customize: the tempo and the mood.
5 Gemini Prompt Mistakes That Kill Your Output
Most weak Gemini answers trace back to a missing instruction rather than a weak model. Adding an explicit format and a length floor fixes a surprising share of them. Instead of "explain this topic", write "explain this topic in at least 200 words using a numbered breakdown".
Here are the five patterns that tend to produce poor output.
No intent clarity. Without a stated output, you get a generic overview. "Tell me about content marketing" returns a textbook summary. "Give me 5 content marketing tactics a B2B SaaS company can run this quarter" returns something you can act on.
Missing format instruction. Gemini writes prose by default. If you want bullets, a table or a numbered list, say so. The model will not guess the right format for your situation.
Bundling unrelated tasks. Two or three different jobs in one prompt tends to dilute the output on all of them. A separate prompt per task takes thirty extra seconds and returns noticeably better work.
No audience. Without one, Gemini writes for everyone. "Explain cloud storage" reads very differently once you add "to a 60-year-old small business owner who has never used it before". Audience shapes tone, vocabulary and depth.
Vague limits. "Keep it short" gives the model room to interpret. "Maximum 150 words" does not. The same applies to "don't make it too technical" against "avoid industry jargon, write for a general audience".
When an answer comes back wrong, the fix is usually one line. Use this table to find it.
Problem | Why it happens | Better instruction |
|---|---|---|
Answer is generic | No audience or purpose given | Add "for [audience], who needs to [do what]" |
Wrong format | Format was never requested | Add "output as a table / bullets / JSON" |
Too much detail | Several jobs in one prompt | Split into separate prompts, one job each |
Tone keeps drifting | No example to match | Paste one sample and say "match this style" |
Facts look invented | No source attached | Upload or quote the source and add "use only this" |
Answer is too short | No length floor set | Add "at least [N] words" and name the sections |
The fix for all five mistakes is the same: go back to the Intent + Context + Constraints formula. For deeper prompting strategy beyond these basics, see our guide on prompt optimization techniques.
Zero-Shot vs. Few-Shot Prompting in Gemini, When to Use Each
Zero-shot prompting means giving Gemini a task with no examples. You describe the intent, context and constraints, and nothing else.
This approach works well for:
brainstorming
first drafts
exploratory tasks
research summaries
It lets the model reason freely without being pulled toward a predefined style.
Zero-shot example: "Write a product description for a wireless ergonomic keyboard. Target audience: remote workers. Tone: conversational but professional. Maximum 80 words. Focus on comfort and productivity benefits."
Few-shot prompting means including one or more examples of the output style you want before giving the task.
This works best when tone consistency or formatting matters, such as:
client deliverables
branded content
recurring content formats
structured responses
Two or three examples are usually enough to guide the model.
Few-shot example: "Here are two examples of how we write product descriptions: [Example 1] [Example 2] Now write a product description for [new product] in the same style."
💡 Placement tip. Put your examples before the task instruction. Gemini identifies the pattern first, then generates the response using the same structure.
How Gemini's Unique Features Change Your Prompting Approach
Gemini connects to the Google tools you already use, which changes what a prompt needs to say. Generic prompt guides treat it as one more chatbot and skip this entirely.
Prompting Gemini with Google Workspace Context
Gemini can pull live data directly from your Gmail, Google Docs, Drive, Sheets and Calendar, so you can reference material instead of pasting it.
Three habits work well inside Workspace:
Reference documents by name or project context rather than describing them from scratch.
Ask Gemini to cross-reference more than one source in a single prompt.
Name the output you need from that material: a summary, a task list, an email draft.
A practical example: "Using the Q3 report in my Drive and the client email thread in Gmail, draft a status update for the exec team. Keep it under 200 words, professional tone, focus on progress and next steps."
That single prompt pulls from two live sources and produces something you can send.
Prompting Gemini's Deep Research Mode
Deep Research is included with Google AI Pro at $19.99 per month, and it works differently from a standard Gemini reply. It searches across many web sources, evaluates them, and compiles a structured report instead of answering in a few seconds.
Pricing is accurate as of the date this article was written and is subject to change. Confirm the current price on the tool's official pricing page before purchasing.
That makes it useful for market research, competitor analysis, industry trend reports and technical briefings. Because it returns a full report, your prompt has to do more work up front.
Deep Research report template
Best for: a structured, sourced report rather than a short answer.
Research [topic]. I need a [word count] report structured as: Executive Summary, Key Findings, Sector Analysis, Risks, Recommendations. Cite all sources. Target audience: [role].
What to customize: the five section names and the audience.
Define the topic, the structure and the audience, and Deep Research has what it needs to stay on track.
How Long Should a Gemini Prompt Be?
A Gemini prompt should be as long as the task needs and no longer, because structure beats length every time. Gemini supports up to 1 million tokens of context, but that capacity is there for your source material rather than your instructions.
Here is a practical guide:
Simple tasks (summary, caption, rewrite): 1-3 sentences is enough.
Professional outputs (emails, reports, strategy docs): 4-7 sentences with explicit format instructions.
Complex Gemini 3.1 Pro tasks: longer structured prompts work best. Include the section headers you want in the response so the model organizes its output to match.
A 3-sentence prompt with clear intent, context and constraints will outperform a 10-sentence prompt that rambles.
Generate Your Own Gemini Prompts, Without Starting From Scratch
A template written for a generic product manager will never match your goal, your audience and your exact output. Phrasly's AI Prompt Generator closes that gap by building the prompt around the task you describe.
You describe what you are trying to do, add a little context, and choose the AI tool you want to prompt. Here is what that looks like on a real task.
✍️ What you type: "I need to tell my client the project will be two weeks late."
📤 What comes back: "Write a client email explaining a two-week delay on [project name]. Explain the cause in one sentence, give the new delivery date, and state what you are doing to prevent further slippage. Warm and accountable tone, under 150 words, subject line included. Do not apologize more than once."
🎯 What the structure added: an audience, a format, a word limit, a tone, and one exclusion. Those five additions are the difference between a usable draft and a generic one.
No prompt engineering experience needed.
Works across the current Gemini models, plus ChatGPT, Claude and other AI tools.
Free to try. No credit card required. An account may be needed for continued use.
If you want the mechanics rather than the tool itself, our guide on how the Gemini prompt generator works walks through it with examples.
Start with one prompt from the quick-pick block above and change only the brackets. Run it, look at what came back, and add the single instruction that was missing: an audience, a format, or a number. That one edit is what separates the prompts people give up on from the ones they keep.
Describe your task once and get a structured Gemini prompt built around it.
What prompt should I use for Gemini?
State the task, the context, the output format and one limit. For example: "Draft a 200-word product update for existing customers, explain what changed and what they must do, use one bulleted list." Change the task and the audience and the shape still holds.
Which Gemini model should I use?
Gemini 3.8 Flash for fast agentic and coding work, Gemini 3.1 Pro for deep reasoning and long uploads. For images use Nano Banana 2, for video Veo 3.1, for music Lyria 3.5. Availability depends on your plan and country.
Can I copy and paste prompts into Gemini?
Yes. Every prompt here is written for it. Replace whatever appears in [square brackets] with your own details, then paste the whole thing into the Gemini chat box. Leave the wording outside the brackets alone on your first run.
How do I add permanent instructions to Gemini?
Open the Gemini app, go to Personal Context, and use the Instructions for Gemini field. Set your default tone, format and length once, and Gemini applies them to every future chat. You can switch the instructions off at any time.
What is the difference between Gemini 3 and Gemini 3.1 Pro?
Gemini 3 is the general model family for everyday tasks. Gemini 3.1 Pro handles complex reasoning, long-context work and multi-step tasks, and it remains the reasoning flagship because Gemini 3.5 Pro was announced and never released.
What is Gemini 3.8 Flash, and how is it different from Gemini 3.1 Pro?
Gemini 3.8 Flash launched on September 2, 2026 and handles long-horizon coding, autonomous agents and multi-step reasoning. Google calls it its most intelligent workhorse model. Use Gemini 3.1 Pro instead when the work needs careful reasoning over long documents.
Can I use the same prompts for Gemini and ChatGPT?
Mostly, yes. Basic prompts transfer well, though Gemini responds better to the Intent + Context + Constraints structure. Workspace access, Deep Research and image generation need Gemini-specific wording. For ChatGPT, see our ChatGPT prompts for professionals guide.
Can Gemini AI generate music with Lyria?
Yes. Lyria 3.5 reached the Gemini app on September 4, 2026, one day after entering public preview in the Gemini API. It also runs through Google AI Studio, Google Flow Music and Google Vids. It generates full tracks with verses and choruses, each carrying a SynthID watermark.
How do I write video prompts for Gemini using Veo?
Name five things: scene, motion direction, camera angle, duration and style. A working format is "Create an 8-second video of [scene]. Motion: [push-in]. Camera angle: [wide]. Style: [cinematic]." Veo runs through the developer tools rather than the chat box.
Can Nano Banana 2 keep my face or subject consistent when editing photos?
Yes. Upload a photo and ask for targeted edits while preserving the same person, pose or product. Say exactly what should stay the same, for example: "Keep the face, pose and outfit unchanged, but replace the background with a modern office."
Is it safe to upload photos to Gemini AI?
For everyday use, yes, with limits. Avoid uploading IDs, confidential work files or sensitive personal photos until you have checked the privacy settings on your account. Apply the same caution you would with any cloud-based tool.
Where can I get free prompts for Gemini?
The 100+ library on this page is free to copy, and Phrasly's AI Prompt Generator builds a custom one around your task. It is free to try with no credit card, though an account may be needed for continued use. Our Claude prompting guide covers other tools.
Is there a free Gemini prompt generator?
Yes. Phrasly's AI Prompt Generator is free to try and builds a structured Gemini prompt around your task, audience and output format. No credit card is required, and an account may be needed for continued use.
Written by
Obaid Ahsan
Content Lead · Pakistan
Obaid Ahsan is a Content Lead with 5+ years of experience in SEO, content strategy, and AI-focused publishing. At Phrasly, he leads content development with a focus on accurate research, search performance, and useful, trustworthy information for readers.
