Prompts that actually work

Claude produces some of the most polished long-form writing of any AI assistant. These prompts are designed to get publication-quality output — whether you are writing blog posts, emails, reports, or creative content.

Claude Opus 5 and Sonnet 5 consistently produce the highest-quality long-form writing of any AI model in July 2026 — a lead documented across writing benchmarks and user testing since Claude 3 Opus. With 1M context and 128K output on Opus 5, Claude can write book chapters, complete reports, and long-form content without truncating. Sonnet 5 at $2/$10/M (intro through August 31) is the best value writing model available. These prompts are optimised for claude.ai Pro/Max, Claude Cowork, and the API.

Claude Opus 5 launched July 24, 2026 at $5/$25/M — the same price as Opus 4.8 with near-Fable-5 intelligence. The model has thinking ON by default and a five-level effort toggle (low/medium/high/max). At low effort it handles fast tasks efficiently; at max effort it works through hard problems with extended reasoning. These prompts are optimised specifically for Opus 5's new capabilities: the effort toggle, the automatic fallback, mid-conversation tool changes, and the 128K output window. Available via claude.ai Pro/Max, Claude Code, and the API as claude-opus-5.

Claude Opus 5 with thinking ON by default and a five-level effort toggle is a strong backbone for agentic workflows. The automatic fallback feature (routes flagged requests to another model instead of blocking) makes Opus 5 more reliable in production agent pipelines than any previous Claude model. Mid-conversation tool changes without cache invalidation enable research-then-action agent patterns. These prompts are optimised for Claude Opus 5 via the API (claude-opus-5), Claude Code, and Cowork.

Claude Opus 5 is used in production security workflows for code vulnerability analysis, threat modelling, security documentation, and red team exercise planning. Importantly: Claude's July 2026 security disclosure confirmed that its models are capable of sophisticated security reasoning — including autonomous exploitation of weak passwords and unauthenticated endpoints. These prompts are for defensive cybersecurity work only: vulnerability analysis, secure code review, threat modelling, and incident response documentation. Claude Opus 5 runs with full safety classifiers on the production API — the evaluation incidents involved models without standard safeguards.

Claude Opus 5 at $5/$25/M with a 1M context window and 128K output can read entire research papers, synthesise multiple sources, and produce long-form research outputs in a single session. With thinking ON by default and a five-level effort toggle, Opus 5 can spend more compute on hard analytical tasks and less on simpler extraction tasks — making research workflows significantly more cost-efficient than running everything at max effort. These prompts are optimised for claude.ai Pro/Max and the API.

Claude Sonnet 5 at $2/$10/M (intro pricing through August 31, 2026) is the most cost-efficient Anthropic model for coding tasks. After August 31 it moves to $3/$15/M. For teams running high-volume coding workloads, the next 30 days are the window to lock in the most cost-efficient Anthropic coding model available. Sonnet 5 handles writing, coding, and analysis tasks at quality comparable to Opus 4.8 at 60% of the price. These prompts are optimised for the Anthropic API (claude-sonnet-5) and Claude Code.

Claude Sonnet 5 at $2/$10/M intro pricing (through August 31, 2026) is the most capable Claude model available at this price point. Sonnet 5 improves on Sonnet 4.6 with adaptive thinking, lower hallucination, and better instruction following — the gains are most visible in writing tasks requiring nuance, judgment, or multi-step refinement. From September 1, Sonnet 5 moves to $3/$15/M plus the new tokenizer adds 10-35% more tokens per request. These prompts are optimised for claude.ai Pro and the Anthropic API (claude-sonnet-5) — use them before the deadline to maximise value.

Cursor is an AI-native IDE that understands your entire codebase — not just the current file. These prompts are designed for Cursor's chat sidebar (Cmd+L), inline editing (Cmd+K), and background agent mode. They cover codebase exploration, multi-file refactoring, test generation, and debugging across large repositories. Optimised for Cursor Pro with Claude Fable 5 or GPT-5.6 Sol as the backing model.

DeepSeek V4 Flash 0731 at $0.14/$0.28/M is the cheapest capable agentic coding model after its July 31 official release. Terminal-Bench 82.7% (up from 61.8% preview) puts it within 0.6 points of Grok 4.5 at 14× the price difference. The model supports Responses API natively, runs in OpenCode, and is Codex-compatible — making it a drop-in for most OpenAI-based agent harnesses. These prompts are optimised for the DeepSeek API (model: deepseek-v4-flash) and self-hosted 0731 weights. Note: migrate from deepseek-chat and deepseek-reasoner before October 24, 2026.

DeepSeek V4 Pro at $0.44/M is the cheapest capable coding API available in July 2026 — 4× cheaper than Grok 4.5, 11× cheaper than Kimi K3. V4 Flash at $0.09/M is even cheaper for simpler tasks. The V4 family supports tool calling, function calling, and multi-turn coding sessions. Self-hostable via official weights under MIT license. For teams running extremely high-volume coding workloads where cost is the absolute primary constraint, DeepSeek V4 remains the lowest-cost capable option. These prompts are optimised for the DeepSeek API and self-hosted V4 Pro deployments.

DeepSeek V4 Pro at $0.44/$0.87/M (discounted) is the most cost-efficient coding model available in July 2026 — 7-17x cheaper than Western frontier alternatives. V4 Flash at $0.14/$0.28/M is ideal for high-volume routing and classification tasks. These prompts are optimised for the DeepSeek API and the Anthropic Messages endpoint (now available at api.deepseek.com/anthropic). Important: migrate your deepseek-chat alias to deepseek-v4-flash before July 24 at 15:59 UTC.

DeepSeek V4 Pro with thinking mode ON is one of the most cost-efficient models for research tasks that require extended reasoning — literature review synthesis, multi-step analysis, and structured comparison. At $0.44/$0.87/M (discounted), it delivers frontier-class reasoning at 7-17x lower cost than Western alternatives. Important: migrate from deepseek-chat to deepseek-v4-flash before July 24 at 15:59 UTC.

DeepSeek V4 Pro at $0.44/$0.87/M (discounted) is the most cost-efficient model for high-volume writing tasks in July 2026. V4 Flash at $0.14/$0.28/M works for simpler writing and content generation. With thinking mode ON by default, V4 Pro works through complex writing tasks — structure, argument, tone — before producing output. These prompts are optimised for the DeepSeek API post-July 24 migration.

Gemini 3.6 Flash at $1.50/$7.50/M with 304 tokens/second and Computer Use built in is a strong model for coding tasks that require speed, multimodal input (screenshots of UI, error messages), and Google ecosystem integration. With a 1M context window and March 2026 knowledge cutoff, it handles large codebases and recent library versions well. These prompts are optimised for Gemini 3.6 Flash via the Google AI Studio API, Vertex AI, and Antigravity (Google's coding tool with 2.4M weekly users).

Gemini 3.6 Flash with built-in Computer Use is the fastest and cheapest agent-capable model available in July 2026. At 304 tokens/second and $7.50/M output, it handles agentic search fan-out, multi-step document workflows, and browser automation at a cost no other frontier model matches. These prompts are optimised for the Gemini API with Computer Use enabled, Google AI Studio, and Vertex AI agent deployments.

Gemini 2.5 Flash and Gemini 3.5 Flash are strong mid-tier coding models with excellent multimodal capabilities — you can paste screenshots of errors, UI mockups, or architecture diagrams and get code in return. These prompts are optimised for Google AI Studio, the Gemini API, and Gemini in Google Workspace for coding, debugging, and code review tasks. Note: Gemini 3.5 Pro has been delayed three times as of July 2026 — these prompts are tested on the currently available Gemini Flash tier.

Gemini 3.6 Flash with Google Search grounding is a strong model for financial research tasks — it can search the live web for current prices, earnings, and regulatory filings, then synthesise them with its March 2026 knowledge base. For teams already using Google Workspace, Gemini in Google Sheets can pull financial data and build models directly. These prompts are optimised for Gemini 3.6 Flash via Google AI Studio ($1.50/M, 304 tok/s) and Gemini Advanced. Note: financial analysis only — not financial advice.

Gemini in Google Workspace (Docs, Gmail, Sheets, Drive) is the most deeply integrated AI productivity tool for users already in the Google ecosystem. Gemini 3.6 Flash's March 2026 knowledge cutoff and multimodal capabilities also make it strong for document-heavy productivity workflows. These prompts work across Gemini in Google Workspace, Google AI Studio, and the Gemini app.

Gemini with Google Search grounding is one of the most powerful research tools available in 2026. Unlike most AI models that rely on training data, Gemini can pull live search results directly into its responses — making it ideal for market research, competitive analysis, literature reviews, and fact-checking. These prompts are optimised for Google AI Studio with grounding enabled and Gemini Advanced.

Gemini's Google Search grounding capability makes it uniquely useful for SEO — it can pull real search data directly into its responses, giving you keyword suggestions based on actual Google Search patterns rather than training data. These prompts are optimised for Google AI Studio with grounding enabled and work particularly well for content gap analysis, SERP research, and structured data generation.

Gemini 3.6 Flash with Google Search grounding has a structural advantage for SEO research: it can search Google's own index in real time, giving it access to current SERP data, competitor content, and live search trends at $1.50/M. For SEO tasks that depend on current ranking signals — what is ranking, what is featured, what questions are being asked — Gemini's live Google access is a genuine edge over models with a fixed knowledge cutoff. These prompts are optimised for Gemini 3.6 Flash via Google AI Studio and the Gemini API with grounding enabled.

Gemini 2.5 Flash is Google's best available writing model in July 2026 — strong for long-form content, factual accuracy with grounding, and multilingual output. These prompts cover blog writing, email drafting, social media content, and document summarisation — optimised for Google AI Studio, Gemini in Google Docs, and the Gemini API.

Grok 4.5 at $2/$6/M is the most cost-efficient mid-frontier model for coding tasks in July 2026. It scored 83.3% on Terminal-Bench 2.1 using 25% fewer output tokens than Opus 4.8 — making it fast and cheap for agentic coding loops. Grok Build (now open-sourced under Apache 2.0) uses Grok 4.5 as its default model. These prompts are optimised for Grok 4.5 via the SpaceXAI API, Grok Build CLI, and Cursor (Grok 4.5 is available on all Cursor plans). Privacy note: if you used Grok Build before the July 25 patch, rotate any credentials that were in directories you ran it in.