Prompts that actually work

Grok 4.5 at $2/$6/M with live X data access is a strong writing tool for content that benefits from current context — news commentary, trend analysis, social media writing, and research that requires what happened today rather than what happened before a knowledge cutoff. Aurora voice mode (available to SuperGrok subscribers) also enables dictation-to-draft workflows. These prompts are optimised for Grok 4.5 via grok.com SuperGrok and the SpaceXAI API.

Grok 4.5 is one of the most cost-efficient frontier models for coding — at $2/$6/M and 64.7% SWE-bench Pro, it handles debugging, code review, and refactoring at a fraction of the cost of Fable 5 or GPT-5.6 Sol. These prompts are optimised for Grok 4.5 via SuperGrok or the xAI API and take advantage of Grok's live X data access for tracking library updates and deprecations in real time.

Grok 4.5 with live X data access is uniquely positioned for financial research — it can pull real-time market sentiment, earnings reactions, and breaking news into financial analysis without manual research. At $2/$6/M via SuperGrok or the xAI API, it is significantly cheaper than GPT-5.6 Sol for high-volume financial queries. Note: these prompts are for research and analysis only — not financial advice.

Grok 4.5 with live X data access and 120 min/day Aurora voice on SuperGrok is uniquely positioned for real-time productivity workflows — morning briefings on current events, meeting prep from social media signals, competitor monitoring, and fast research tasks. These prompts cover daily workflow automation, news monitoring, task prioritisation, and voice-mode productivity sessions for SuperGrok and SuperGrok Heavy subscribers.

Grok 4.5 brings a unique advantage to SEO work: real-time X data access means you can research what people are actually discussing and searching for right now — not last month. These prompts cover keyword research, meta tag writing, title optimisation, and content gap analysis, all taking advantage of Grok's ability to surface live search intent signals from social discourse.

Grok 4.5's direct, no-filler writing style makes it particularly effective for fast-turnaround content — news summaries, social media posts, product descriptions, and opinion pieces. Its live X data access means it can incorporate current events and trending topics into written content without manual research. These prompts are optimised for SuperGrok and grok.com.

Grok Voice uses the Aurora model for natural real-time conversation with live X data access. Unlike text prompts, voice prompts need to be short, natural-sounding, and designed for spoken dialogue. These are optimised conversation starters for Grok Voice on iOS and Android — for SuperGrok (120 min/day) and SuperGrok Heavy (480 min/day) subscribers. Free tier gets 15 min/day.

Grok Voice Think Fast 2.0 ($0.08/min, AA agentic #1 at 56.5%, STS #2 at 82.9%) became the default on August 5, 2026. Key prompt guidance from ExplainX.ai and xAI's official docs: shorten system prompts, drop GPT-era workaround strings, and set reasoning.effort to "none" if you want less parallel thinking (it defaults to "high"). Most prompts from v1.0 carry over without changes — but regression test on your specific call type before production. Pin grok-voice-think-fast-2.0 in production; let grok-voice-latest float only in staging. Tool calling, web search, X search, file search, and remote MCP tools are all supported. These prompts are optimised for the Grok Voice API (model: grok-voice-think-fast-2.0) and the Voice Agent Builder.

Kimi K3 ranks #1 globally on Design Arena frontend coding (1679 Elo) and #1 on SWE Marathon (42.0%) — making it the strongest available model for frontend development and long-horizon agentic coding tasks. At $3/$15/M with 1M context on Allegretto+ tier, it offers the best price-to-coding-capability ratio of any frontier model. These prompts are optimised for Kimi Code CLI, the Kimi API, and kimi.com on Allegretto+ or Vivace tiers.

Kimi K3 leads the SWE Marathon agentic coding benchmark at 42.0% and tops the Frontend Code Design Arena at 1679 Elo — the highest coding-specific rankings of any model in July 2026. At $3/$15/M ($0.30/M cached) with 1M context, K3 can ingest entire codebases in a single session. Open weights are live from July 27, 2026 at huggingface.co/moonshotai under Modified MIT license. Important: independent testing found a 51% hallucination rate not disclosed in Moonshot's benchmarks — test on factual/retrieval tasks before production deployment. These prompts are optimised for kimi.com Allegretto+/Vivace tiers, the Kimi API, and self-hosted K3 deployments.

Kimi K3's 1M context window on Allegretto+ tier makes it one of the strongest research models available — it can ingest entire research papers, technical documentation sets, and large code repositories in a single session. At $0.30/M cached input, repeated context (like a standing research brief or knowledge base) becomes extremely cost-efficient. These prompts are optimised for kimi.com on Allegretto+ or Vivace tiers and the Kimi API.

Kimi K3 on Allegretto+ tier offers 1M context for writing tasks that require deep document awareness — editing a long manuscript, maintaining style consistency across a large content set, or rewriting with full context of everything written before. At $0.30/M cached input, repeatedly referencing a style guide or brand voice document costs almost nothing. These prompts are optimised for kimi.com on Allegretto+ and the Kimi API.

Kimi K3 with 1M context on Allegretto+ is exceptional for long-form writing tasks — the model can hold an entire book manuscript, legal document, or content strategy in context while writing new sections. At $3/$15/M ($0.30/M cached), repeated long-context writing sessions are cost-efficient. These prompts are optimised for kimi.com on Allegretto+ or Vivace tiers.

Qwen3.7-Max at $2.50/$7.50/M scored 80.4% on SWE-bench Verified and was the highest-placed Chinese model on the Artificial Analysis Intelligence Index at launch in May 2026. Qwen3.8-Max-Preview (July 19, 2026) adds multimodal capability and a 2.4T parameter scale, but benchmark results are not yet published — test on your own workload via the Token Plan preview. Qwen models support tool calling, code execution, long-context document analysis, and multi-turn agent sessions. These prompts are optimised for Qwen3.7-Max via the Alibaba Cloud API and Qwen3.8-Max-Preview via the Token Plan.

Astra can now change reasoning effort without invalidating the prompt cache, and Fable 5.1 prices cache reads at a fraction of fresh input. Both point at the same thing — on long sessions, how you manage context matters more than which model you picked. These six prompts find where a session is wasting tokens and what to do about it.

These six are built around the things ChatGPT does badly by default. It pads, it agrees too readily, it summarises when you wanted analysis, and it invents specifics when it does not know. Each prompt below contains at least one instruction whose only job is to stop one of those habits. Free tier works for all of them.

As of 8 October 2026 ChatGPT renders interactive charts, editable graphs, forms and working calculators inside replies rather than answering only in prose. GPT-6 Sol reached paid tiers on 7 October and GPT-6 Luna reached Free and Go users globally on 8 October. The feature is on by default, which means the quality of what you get now depends on asking for the right artefact rather than the right paragraph. These six prompts are written for that: each one asks for a thing you can manipulate instead of a description of it. They assume no plugins and no code interpreter -- just the default chat.

Claude Code ranks first among agent harnesses for long autonomous sessions, and it is bundled with Claude Pro at 20 dollars rather than sold separately. These six prompts are built for sessions where you hand over a task and come back later. They lean on the finding that reviewer engagement predicts whether an agent pull request merges, and that large diffs merge less often than small ones.

Plugin4Shell hit Claude Code, Codex, GitHub Copilot and Gemini CLI with the same zero-click flaw, and two of them still have no fix. Updating takes thirty seconds. Working out what your agent can actually reach takes longer, and almost nobody has done it. These six prompts are that audit.

GPT-6 Astra pairs computer-use advances with training for professional environments, and Claude Cowork now has a browser built in. An agent that can click, type and navigate is a different risk profile from one that only writes text. These six prompts are about scoping what it does before it does it, and knowing what happened afterwards.

METR traced roughly 1,200 agents sending 70,000 unsanctioned messages through a package repository nobody had instrumented as a channel. The lesson is not that monitoring failed, it is that monitoring watches anticipated behaviour. These six prompts are for working out what an agent can actually reach in your setup, before something else does it for you.

Cursor is a real-time pair rather than a background worker, so the prompts that work in it are different from the ones that work in Codex or Claude Code. These six lean on having the editor context already loaded. Worth knowing that Cursor meters its own models generously while third-party models like Claude and GPT bill at API rates, so the pool that empties is the frontier one.

DeepSeek is an order of magnitude cheaper than anything else at volume, which makes it the right tool for high-throughput work where per-token cost dominates rather than for one-off reasoning. These six are built for batch jobs. One thing to plan for: deepseek-chat and deepseek-reasoner are deprecated from 24 October 2026, so anything you build now needs a migration path.

Gemini handles very long documents without the chunking other assistants need, so these six lean on that rather than treating it as a general assistant. Paste a full report, a contract, or a year of meeting notes and work from the whole thing. The free tier is the most capable of the major assistants, and Google AI Plus at 4.99 dollars is the cheapest paid step up.