AI & Prompt Tools · Free tool
AI Prompt Library
Browse a curated catalog of prompt templates for writing, coding, marketing, and research. One click to copy.
Rewrite the following text to be 30% shorter while keeping every concrete fact, number, and name. Remove adjectives, hedging, and repetition. Do not add new ideas. Return only the rewrite.
Text:
{paste here}Turn this single bullet into one tight paragraph (3–5 sentences). Keep the voice matter-of-fact. No intro phrases like 'Here is…'
Bullet:
- {paste here}Write a 4-sentence email declining the request below. Be warm, specific about why not, and offer one concrete alternative.
Situation:
{paste here}Rewrite this paragraph so a smart reader outside the field can follow it. Define the first use of each specialized term in 5 words or fewer. Keep sentences short.
Text:
{paste here}Create an H1, three H2 sections, and 2-3 H3 bullets per section for a blog post about: {topic}. Audience: {reader}. Angle: {angle}. Each H3 should hint at a specific example, not a generic subheading.Explain this git diff. For each changed hunk: (1) what the change does in plain English, (2) why someone might make this change, (3) one thing that could go wrong.
```diff
{paste diff}
```Write unit tests (framework: {jest/vitest/pytest}) for the function below. Cover: happy path, empty input, boundary values, error conditions. Each test should have a descriptive name.
```
{paste function}
```Here is an error message and the smallest snippet of code where it happens. Explain the root cause in 2–3 sentences, then show the corrected snippet with only the needed changes.
Error:
{paste}
Code:
```
{paste}
```Write a regex that matches {describe}. Include: (1) the regex itself, (2) a one-line explanation, (3) three matching examples and three non-matching examples.Refactor this function for readability, not performance. Name things well, remove nested conditionals, and keep behavior identical. Output only the refactored code plus a 2-line summary of what changed.
```
{paste}
```Write 10 headlines for: {product or article}. Mix these styles: direct benefit, curiosity gap, contrarian, numbered, how-to, and social proof. No clickbait. Each under 70 characters.Write a 90-word cold email. Recipient: {role at company}. Hook: {their recent thing}. Offer: {what I bring}. Ask: one 15-minute call. No 'hope this finds you well'. First line must name something specific and true about them.Write 3 hero-section variants for the landing page of {product}. Each variant: 1 headline (≤10 words), 1 subhead (≤25 words), 1 CTA button text (≤3 words). Avoid hype words.Turn this article into: (a) 1 LinkedIn post (~150 words, first-person, one concrete example), (b) 1 Twitter/X thread of 5–7 tweets. No emojis unless absolutely needed.
Article:
{paste}I have 5 minutes to understand {topic}. Explain: (1) what problem this solves, (2) the main approaches and how they differ, (3) the sharpest open question. Assume I'm smart but have no background in this field.Compare {X} and {Y} across: typical use cases, cost, learning curve, ceiling of capability, and one-thing-each-is-bad-at. Return a markdown table. No hedging — pick a clear winner per row when one exists.I believe: {my position}. Steelman the strongest counter-position. Present 3 arguments a thoughtful critic would make, and for each: one piece of evidence and one concession I should make if I still disagree.Rewrite this resume bullet using the format 'Action verb + what + measurable impact'. Remove buzzwords. Keep under 20 words.
Bullet:
{paste}I'm prepping for the question: '{question}'. Draft a 90-second STAR-format answer based on this situation: {context}. End with the lesson learned in one sentence.Write a negotiation response. Offer I got: {offer}. My target: {target}. My leverage: {reason}. Keep it warm, firm, and under 100 words. End with a clear ask, not a question.Explain {concept} as if I'm 12. One analogy, one real-world example, and one common misconception — in that order. Max 150 words.Turn the notes below into 10 Anki-style Q/A flashcards. Make each question specific enough that guessing is impossible. Front: question. Back: answer + one-line 'why it matters'.
Notes:
{paste}Act as a Socratic tutor on {topic}. Don't explain directly — ask me questions one at a time to guide me toward understanding. Start with the most basic one. Wait for my answer before moving on.From these raw notes, extract: (1) decisions made, (2) action items in the format 'Owner — action — due date', (3) open questions with nobody assigned. Skip small talk.
Notes:
{paste}Help me build an agenda for a 20-minute conversation with {person}. The hard topic: {topic}. Give me: (1) opening line, (2) 3 questions to understand their side, (3) the one truth I need to say, (4) a good exit line.Advertisement
What it does
Browse a curated catalog of prompt templates for writing, coding, marketing, and research. One click to copy. Production AI usage typically costs 3-5x what initial estimates suggest because of output-token weighting and prompt-cache misses.
Token counting matters: input vs output tokens, system-prompt overhead, RAG context all affect bills. The gap between “rough estimate” and “defensible number” is exactly where good tooling earns its keep — the math is reproducible, but knowing which inputs matter and what the result means is half the work.
Prompt caching (OpenAI: 50% off cached tokens; Anthropic: 90% off) is the single biggest cost optimization for chatty workloads. A common pitfall: treating model output as authoritative without verification. Treat the tool’s output as a starting point and validate against authoritative sources for any consequential decision.
Embed this tool on your siteShow snippetHide
Paste this snippet into any page. Loads on-demand (lazy), no tracking scripts, and sized to most dashboards. Replace the height to fit your layout.
<iframe src="https://freetoolarena.com/embed/ai-prompt-library" width="100%" height="720" frameborder="0" loading="lazy" title="AI Prompt Library" style="border:1px solid #e2e8f0;border-radius:12px;max-width:720px;"></iframe>How to use it
- Open the tool and review the interface.
- Enter or paste your input.
- Configure any relevant options.
- Run the tool and review the output.
- Iterate or refine based on the result.
When to use this tool
- Comparing API costs vs self-hosting for high-volume workloads.
- Production cost forecasting based on traffic projections.
- Prompt-engineering optimization to reduce token consumption.
- Vendor selection between OpenAI, Anthropic, Google, and open-source.
When not to use it
- When the workload is unique enough that public benchmarks don’t apply.
- For non-frontier image, video, or audio model pricing (those use per-asset billing).
- When you have negotiated enterprise pricing not reflected in public rate cards.
- For hyper-bursty traffic where peak load determines architecture, not average.
Common use cases
- A ML engineers optimizing inference costs working through ai prompt library for a real decision.
- A developers building LLM features working through ai prompt library for a real decision.
- A researchers comparing model quality working through ai prompt library for a real decision.
- A enterprise teams managing AI budgets working through ai prompt library for a real decision.
Frequently asked questions
- How does this compare to GPT-4o or Claude Opus 4?
- GPT-4o, Claude Opus 4, and Gemini 2.5 Pro are roughly comparable on quality for general tasks; their pricing differs by 30-50% so test on your specific workload before locking in.
- What hidden costs am I missing?
- Output tokens (3-5x input cost), rate-limit retry overhead (20-40% extra), failed-request charges, and the engineering time to maintain the integration. Budget 1.5-2x the headline rate.
- How does self-hosting change the math?
- Self-hosting Llama 3.3 70B on AWS p4d ($32/hr) costs ~$16/M tokens at full utilization. DeepSeek V3 API is $0.30/M tokens. Self-hosting wins only at 1B+ tokens/month consistent.
- Should I switch to a smaller model?
- Probably yes, after testing. Mini / Haiku tier handles 60-70% of production tasks adequately at 5-10x lower cost. Test on your specific workload, then route only failures to the larger model.
- What about prompt caching and batch discounts?
- Prompt caching saves 50-90% on cached input tokens (OpenAI: 50%; Anthropic: up to 90% with 5-minute cache). Batch API: 50% off async jobs. Combined, can drop bills 70-80% for cache-friendly workloads.
- Is this calculation accurate at scale?
- Public-rate-card calculators are accurate within 10-15% for typical workloads. Variance comes from prompt-cache hit rates, batch-API usage, and rate-limit retry overhead.
Advertisement
Learn more
Guides about this topic
- AI & LLMs · GuideHow to Set Up an AI AgentNavigate a plain-English decision tree to pick the right AI agent stack for 2026. Free, instant online walkthrough, no sign-up.
- AI & LLMs · GuideHow to Use ChatGPT Agent ModeWhere /agent is available (Plus, Pro, Team — not Free), the 8 tasks it actually does well, and the 5 it can't. Plus the briefing template that works.
- AI & LLMs · GuideHow to Build an Agent with the OpenAI Agents SDKBuild a working Python agent with OpenAI's Agents SDK — tools, handoffs, guardrails, and the model-native sandbox harness. Free guide, no sign-up needed.
- AI & LLMs · GuideHow to Build an Agent with the Claude Agent SDKBuild an agent with the Claude Agent SDK — install, write custom tools, add hooks, compose sub-agents on the harness powering Claude Code. Free guide.
- AI & LLMs · GuideHow to Set Up Claude CodeConfigure Claude Code with permissions, MCP servers, and sub-agents for a full working setup. Free browser-only guide, no sign-up.
- AI & LLMs · GuideHow to Set Up Cursor AI IDEOptimize Cursor AI IDE modes, .cursorrules, and model picks to avoid credit-pricing traps. Free, instant configuration guide, no sign-up.
Explore more ai & prompt tools tools
- AI Image Prompt HelperBuild effective image prompts: pick style, lighting, camera, aspect ratio, extras. Outputs prompt + negative prompt for Midjourney, DALL-E, FLUX, SD 3.5.
- Open-Source LLM TrackerLive tracker of 15 open-weight LLMs: Llama 3.3/4, Qwen 3.5, DeepSeek V3.2/R1, Kimi K2, Mistral Large 3, Gemma 3, Phi-4, SmolLM3. Filter by license.
- AI Transcription Tools Compared9 transcription tools compared: Otter, Whisper API, Deepgram Nova-3, AssemblyAI, Rev, Sonix, Granola, Zoom AI, MacWhisper. Accuracy, languages, pricing.
- AI Data Residency CheckerFind AI providers compliant with your region (US, EU, UK, APAC, Canada) and certifications (SOC 2, HIPAA). Includes Bedrock, Azure, Mistral, self-host.
- AI Context Window PlannerPlan your prompt budget across system + docs + history + output + buffer. See which AI models (Claude, GPT, Gemini, DeepSeek, Kimi) fit your needs.
- AI Agent Platforms Compared10 agentic AI platforms compared: ChatGPT Operator/Atlas, Claude Computer Use, Devin, Manus, Replit Agent, Cursor Background Agents, Bolt.new, v0, Lovable.