{"id":"C01","category":"code","prompt":"Write a Python function that parses a CSV string and returns the sum of a named numeric column. Include a short explanation of how it works and one example call."}
{"id":"C02","category":"code","prompt":"Write a JavaScript function that debounces a function call by a given number of milliseconds. Include a short explanation of how it works and one example call."}
{"id":"C03","category":"code","prompt":"Write a Python function that checks whether two strings are anagrams, ignoring case and spaces. Include a short explanation of how it works and one example call."}
{"id":"C04","category":"code","prompt":"Write a SQL function that returns the second highest salary from an employees table. Include a short explanation of how it works and one example call."}
{"id":"C05","category":"code","prompt":"Write a JavaScript function that flattens a nested array of arbitrary depth without using Array.flat. Include a short explanation of how it works and one example call."}
{"id":"C06","category":"code","prompt":"Write a Python function that merges two sorted lists into one sorted list in linear time. Include a short explanation of how it works and one example call."}
{"id":"C07","category":"code","prompt":"Write a Go function that reads a file line by line and counts the lines that contain a given word. Include a short explanation of how it works and one example call."}
{"id":"C08","category":"code","prompt":"Write a Python function that validates that a string of brackets is balanced. Include a short explanation of how it works and one example call."}
{"id":"C09","category":"code","prompt":"Write a JavaScript function that groups an array of objects by a key and returns a map of arrays. Include a short explanation of how it works and one example call."}
{"id":"C10","category":"code","prompt":"Write a Rust function that reverses the words in a sentence while keeping punctuation in place. Include a short explanation of how it works and one example call."}
{"id":"C11","category":"code","prompt":"Write a Python function that computes the n-th Fibonacci number using memoisation. Include a short explanation of how it works and one example call."}
{"id":"C12","category":"code","prompt":"Write a TypeScript function that defines a generic function that returns the unique items of an array. Include a short explanation of how it works and one example call."}
{"id":"C13","category":"code","prompt":"Write a Python function that finds the longest common prefix of a list of strings. Include a short explanation of how it works and one example call."}
{"id":"C14","category":"code","prompt":"Write a Bash function that renames every .jpeg file in a folder to .jpg. Include a short explanation of how it works and one example call."}
{"id":"C15","category":"code","prompt":"Write a Python function that converts a Roman numeral string to an integer. Include a short explanation of how it works and one example call."}
{"id":"C16","category":"code","prompt":"Write a JavaScript function that implements a simple LRU cache class with get and set. Include a short explanation of how it works and one example call."}
{"id":"C17","category":"code","prompt":"Write a SQL function that finds customers who placed more than three orders in the last 30 days. Include a short explanation of how it works and one example call."}
{"id":"C18","category":"code","prompt":"Write a Python function that implements binary search and returns the index or -1. Include a short explanation of how it works and one example call."}
{"id":"C19","category":"code","prompt":"Write a Java function that checks whether a number is prime and returns a boolean. Include a short explanation of how it works and one example call."}
{"id":"C20","category":"code","prompt":"Write a Python function that removes duplicate rows from a list of dictionaries by a given key. Include a short explanation of how it works and one example call."}
{"id":"C21","category":"code","prompt":"Write a JavaScript function that fetches JSON from a URL with a timeout and one retry. Include a short explanation of how it works and one example call."}
{"id":"C22","category":"code","prompt":"Write a Python function that rotates a matrix 90 degrees clockwise in place. Include a short explanation of how it works and one example call."}
{"id":"C23","category":"code","prompt":"Write a C# function that splits a list into chunks of a given size. Include a short explanation of how it works and one example call."}
{"id":"C24","category":"code","prompt":"Write a Python function that tokenises a sentence into lowercase words and counts word frequencies. Include a short explanation of how it works and one example call."}
{"id":"C25","category":"code","prompt":"Write a Python function that writes a decorator that logs how long a function takes to run. Include a short explanation of how it works and one example call."}
{"id":"C26","category":"code","prompt":"Write a JavaScript function that deep-clones a plain object without JSON methods. Include a short explanation of how it works and one example call."}
{"id":"C27","category":"code","prompt":"Write a SQL function that calculates a running total of sales by day. Include a short explanation of how it works and one example call."}
{"id":"C28","category":"code","prompt":"Write a Python function that finds all pairs in a list that sum to a target value. Include a short explanation of how it works and one example call."}
{"id":"C29","category":"code","prompt":"Write a Go function that implements a worker pool that processes jobs from a channel. Include a short explanation of how it works and one example call."}
{"id":"C30","category":"code","prompt":"Write a Python function that converts a number of seconds into a human readable duration string. Include a short explanation of how it works and one example call."}
{"id":"S01","category":"summarise","source":"articles/best-ai-transcription-tools-2026.html","prompt":"Summarise the following passage in three short bullet points.\n\nAI transcription in 2026 is a buyer's market, but the tool that tops an accuracy benchmark often fails on the decision that actually matters for your team: whether it integrates with Zoom, whether it signs a BAA , or whether it costs $0.006 per minute or $1.99. The short answer: Otter.ai for team meeting notes, Rev when you need human backup, Fireflies.ai for CRM-connected sales calls, Deepgram or AssemblyAI for developer pipelines, and Whisper when you need zero marginal cost and full data control. The comparison matrix below is designed to be the one table you bookmark before making this decision. Raw word error rate is the most commonly cited benchmark, but it is often the least useful differentiator for buyers, because every serious tool cleared the 90 percent accuracy bar on clean audio several years ago. The real separators in 2026 are the decisions that get made after the first 90 percent:"}
{"id":"S02","category":"summarise","source":"articles/perplexity-vs-claude-vs-chatgpt-deep-research-2026.html","prompt":"Summarise the following passage in three short bullet points.\n\nMost \"best AI for deep research\" roundups end the same lazy way: \"no single winner, use all three.\" That is technically true and practically useless. So we did the thing those pages skip: instead of vague blurbs, we scored the four deep-research modes (not the chatbots) on one transparent metric across speed, citation reliability, report depth, free access, and cost. The scores are an editorial composite built from verified public product facts, not a black box, so you can see exactly why the verdict lands where it does and reweight it yourself. If you want our broader pillar on the category first, read our best AI for research guide. The contrarian headline you will not see led on most roundups: even the citation leader fabricates more than one source in three. The DRFI , or Deep-Research Fitness Index, is a single 0-to-100 score we compute for each mode from five weighted dimensions: citation reliability, report depth, speed, free access, and cost. It exists because the usual \"use all three\" non-answer hides the fact that these modes are genuinely good at different jobs, and a one-number summary plus a transparent formula lets you see the trade-off at a glance instead of wading through four vendor brochures."}
{"id":"S03","category":"summarise","source":"articles/tidio-pricing-2026.html","prompt":"Summarise the following passage in three short bullet points.\n\nTidio is worth it when chat volume is modest and the Lyro AI Agent can answer repeat questions, because only human-replied conversations count against the quota. Tidio says its customers average a 75% resolution rate with Lyro, which is the vendor's own figure and not an independent result. Every page carries a labelled sponsored shelf beside our editorial ranking: slots from $1,200 for a 6-month term, each with a visible Sponsored label and a link marked as paid for search engines, and the top of the shelf goes to the highest bid. It never changes what we recommend, how we rank, or what a verdict says . Position is not for sale here at any price, which is the reason a placement beside it is worth having. Your competitors can buy in too; the shelf is ordered by bid."}
{"id":"S04","category":"summarise","source":"articles/best-ai-meeting-notes-for-consultants-and-account-managers-2026.html","prompt":"Summarise the following passage in three short bullet points.\n\nTwelve note takers compared on the four things client work demands: a consent notice before the call, a bot or no bot, a client-safe summary, and notes that land in HubSpot or Salesforce. Every price was read on the vendor’s own page on 19 October 2026. Method: for each vendor, the lowest annual per-seat price on the plan its own pricing page or help center names for a native HubSpot or Salesforce sync, read 19 October 2026. Fathom, Sembly, MeetGeek and tl;dv do not state the plan clearly enough to include. See the table and how this page was built ."}
{"id":"S05","category":"summarise","source":"articles/blink-pricing-2026.html","prompt":"Summarise the following passage in three short bullet points.\n\nPaid plans also get 5 credits a day, and Blink's FAQ says that adds up to 150 credits a month with no monthly cap. If you used every daily credit, Starter's $25 would buy up to 250 credits, about $0.10 each. That is a ceiling, not a typical month. For a side-by-side with Lovable on credit price, yearly price and rollover, see Blink vs Lovable pricing . Blink competes with other AI app builders. Nesyona compares them in best AI app builders , and covers the coding side in best AI coding assistants and the no-code side in best AI website builders ."}
{"id":"S06","category":"summarise","source":"articles/best-ai-chatbots-2026.html","prompt":"Summarise the following passage in three short bullet points.\n\nThe AI chatbot market in 2026 is a four-way race with two dark horses. We tested ChatGPT (GPT-4o), Claude (Opus and Sonnet), Gemini (Ultra and Pro), Perplexity, Microsoft Copilot, and Mistral Le Chat with identical prompts across writing, coding, research, math reasoning, and creative tasks. No chatbot wins everything. Here's which one is best for what you actually need it for. Claude is the chatbot we reach for when the task requires nuance. In our side-by-side testing, Claude Opus produced the most natural-sounding prose, handled complex multi-step reasoning more reliably than GPT-4o, and maintained coherence across 100K+ token conversations without degradation. Sonnet (the mid-tier model) is faster and still outperforms GPT-4o on most coding benchmarks while being free to use with generous limits."}
{"id":"S07","category":"summarise","source":"articles/best-ai-browser-agents-2026.html","prompt":"Summarise the following passage in three short bullet points.\n\nAn AI browser agent is software that drives a web browser to complete a goal you describe in plain language, clicking and typing across pages instead of just answering a question. In 2026 the category finally crossed from demo to daily-driver for narrow tasks, but the gap between the best and worst is enormous on anything multi-step. We ran the same five web tasks through every major agent. The short answer: ChatGPT Agent and Claude computer use are the most reliable general picks, Manus wins long autonomous runs, and the open-source browser-use library wins for developers. Want help assembling the rest of your toolkit around an agent? Run your workflow through our AI stack optimizer in about 30 seconds. Looking for a full AI browser to replace Chrome rather than an automation agent? See ChatGPT Atlas vs Perplexity Comet . An AI browser agent is a system that controls a web browser to complete a goal, taking real actions across pages rather than only generating text. It perceives the page (through screenshots, the accessibility tree, or the raw DOM ), decides on a next action, executes it, observes the result, and repeats until the goal is met or it hits a wall. That perceive-decide-act loop is what separates it from a chatbot, which stops at the answer."}
{"id":"S08","category":"summarise","source":"articles/best-ai-sales-copilots-2026.html","prompt":"Summarise the following passage in three short bullet points.\n\nMost sales AI marketing in 2026 collapses very different products into one shopping aisle. The actual buying decision splits along architectural shape. Conversation-intelligence platforms observe and coach. CRM-embedded copilots sit inside the workflow the rep already uses. Autonomous agents execute outbound. The right pick depends less on which vendor scores highest on a feature matrix and more on which shape fits your sales motion, which CRM you already pay for, and how much of the SDR throughput you want to outsource to software. We scored Gong , Clari Copilot , Salesforce Agentforce , HubSpot Breeze , Outreach (Kaia plus AI) , Apollo.io , 11x (Alice and Mike) , and Artisan AI (Ava) across ten capability axes, mapped pricing transparency, and broke picks down by sales motion and revenue-operations persona. Match a stack to your motion with our AI stack optimizer (sales lane) in 60 seconds, track upcoming pricing changes in the AI tool pricing tracker , or sharpen your discovery prompts in the prompt compiler . Jump to the decision fork . Sales AI in 2026 is not one product category. It is three product categories that the marketing aisle has compressed into one shopping experience. The fastest way to lose a procurement cycle is to compare a conversation-intelligence platform against an autonomous-agent platform on a feature matrix and decide on the wider checklist. The two solve different problems for different teams. Use the vs-grid below to anchor which shape your sales motion actually needs, then drop into the capability matrix to filter vendors."}
{"id":"S09","category":"summarise","source":"articles/cheapest-ai-model-for-your-task-2026.html","prompt":"Summarise the following passage in three short bullet points.\n\nThe top model on the LMArena text leaderboard when we checked, Gemini 4 Argon, lists at $4.00 blended per million tokens. The cheapest model within 50 rating points of it lists at $0.54. This picker lets you set the gap yourself. It is free to embed. Slide to the quality gap you can live with and see the cheapest model that clears it. Today's top model is Gemini 4 Argon at a rating of 1526; the cheapest model within 50 points of it is MiMo-V2.6-Pro at $0.54 blended per million tokens, against the leader's $4.00."}
{"id":"S10","category":"summarise","source":"articles/best-ai-candidate-sourcing-tools-2026.html","prompt":"Summarise the following passage in three short bullet points.\n\nSix sourcing tools compared on the data underneath, the outreach on top, and what a seat really costs. Published entry prices run from $15 to $494 a month, a 33x spread, and the three tools that publish a profile count sit between $99 and $149. Prices read on each vendor’s own page on 19 October 2026. Published entry prices for AI candidate sourcing tools span 33x, from Manatal at $15 a user a month (annual) to hireEZ at $494 a month for a solo recruiter, yet the three tools that publish a profile count (Juicebox, Gem, SeekOut) all sit between $99 and $149: database size does not explain the spread."}
{"id":"S11","category":"summarise","source":"articles/best-ai-logo-brand-identity-generators-2026.html","prompt":"Summarise the following passage in three short bullet points.\n\nAI logo generators have compressed a process that used to cost $500 to $5,000 and take weeks into a $20 to $100 one-time purchase that takes an afternoon. The category is genuinely useful, but the tools optimize for very different buyers, and several come with real limits on vector output, ownership rights, and design differentiation that most review sites gloss over. Short answer: Looka for a complete brand kit, Brandmark for the cleanest one-time-purchase logo, Canva for teams already inside its ecosystem, and Adobe Express for Creative Cloud subscribers. If you would rather get the logo and a full branding suite in one place, Design.com's AI logo generator bundles logo creation with 50+ branding tools on a free-to-start plan. An AI logo generator takes a short input, usually a brand name, industry selection, and style preferences, and produces a set of logo concepts using a combination of trained style models, stock icon libraries, and type rendering. The better tools then let you iterate on the output: swap icons, adjust spacing, change color palettes, and download the result in multiple formats."}
{"id":"S12","category":"summarise","source":"articles/suno-vs-udio.html","prompt":"Summarise the following passage in three short bullet points.\n\nTwo platforms now dominate AI music generation : Suno and Udio . Both turn a text prompt into a full song with vocals, instrumentation, and mixing. Both charge $10 a month for the entry paid tier ($8 billed yearly). They have diverged hard since 2024: Suno is now a full production environment with v6 models, stem separation from the Pro tier and the Suno Studio DAW. Udio still makes cleaner raw audio, but since its October 2025 Universal Music deal nothing you make there can be downloaded. We tested both on the same prompts, same vocals, same genres. Here is what matters when you choose between them. We ran an identical brief through both platforms: \"upbeat indie-pop track, 110 BPM, female vocal, hopeful but introspective, dreamy synth bed, real drums.\" Three takes each. Below is the player view of the best take from each, plus which plan lets you export stems today (September 2026)."}
{"id":"S13","category":"summarise","source":"articles/best-ai-website-builders.html","prompt":"Summarise the following passage in three short bullet points.\n\nAI website builders now produce professional, SEO-ready sites from a single text prompt in under 60 seconds. What took weeks of development and thousands of dollars now costs roughly $10 to $30 a month. Here's which tools actually deliver - and which are just glorified templates. We ran the same prompt through every builder on this list: \"Create a website for a boutique coffee roastery in Portland with an about page, menu, online ordering link, and contact form.\" We evaluated each output on five criteria: generation speed (time from prompt to live preview), design quality (layout, typography, spacing, mobile responsiveness), customization depth (how much you can change after generation), SEO readiness (heading structure, meta tags, page speed), and pricing transparency (what you actually pay after the free trial ends)."}
{"id":"S14","category":"summarise","source":"articles/best-ai-note-taking-apps.html","prompt":"Summarise the following passage in three short bullet points.\n\nThe average professional spends 31 hours per month in meetings. Most of that time produces notes that are incomplete, action items that are forgotten, and decisions that nobody remembers making. AI meeting assistants fix this by joining your calls (Zoom, Teams, Google Meet), transcribing everything, identifying speakers, extracting action items, and generating searchable summaries, all automatically. We tested 8 tools to find which ones actually turn meetings into useful artifacts instead of more noise. Fireflies joins your meetings as a silent bot, transcribes everything with 95%+ accuracy (including speaker identification), and generates structured summaries with action items, key topics, questions asked, and even sentiment analysis (did this meeting feel tense or collaborative?). The AI search lets you query across all your past meetings, \"what did Sarah say about the Q3 budget?\" returns the exact moment with timestamp and context."}
{"id":"S15","category":"summarise","source":"articles/best-ai-writing-tools.html","prompt":"Summarise the following passage in three short bullet points.\n\nThere are dozens of AI writing tools and they all sound the same in their own marketing. We tested the five that matter on the same brief, scored each on quality, voice, templates, and value, and ranked them into a clean three-tier order. The winner is not the most expensive. The most expensive is not the best value. And the cheapest credible option produces the best prose. Here is the tier ladder and the prose each one actually generated. If the job is making writing shorter rather than generating it, the controls matter more than the model: our comparison of AI writing tools for concise output maps which tools offer a shorten action, a length target or a readability grade, and which of those are free."}
{"id":"S16","category":"summarise","source":"articles/best-ai-scheduling-productivity-assistants.html","prompt":"Summarise the following passage in three short bullet points.\n\nWe used each scheduling tool as our primary calendar manager for a full work week - five days with a realistic mix of meetings, focus blocks, recurring tasks, and deadline-driven work. Our test workflow included 8-12 meetings per week, 3 daily habits (exercise, lunch, deep work), 15-20 tasks with varying deadlines, and cross-calendar coordination between a personal Google Calendar and work Outlook account. We measured focus time gained (hours of uninterrupted deep work), scheduling accuracy (did the AI place tasks at sensible times), rescheduling responsiveness (how quickly the system adapted when meetings moved), and setup time (how long from sign-up to a functional workflow). All pricing verified on vendor sites in April 2026. Motion combines a calendar, task manager, and project manager into one interface - and then uses AI to schedule everything automatically. Add a task with a deadline, and Motion finds the optimal time slot based on your energy patterns, existing commitments, and task priority. A meeting runs 15 minutes long? Everything downstream reschedules dynamically. No manual reshuffling required."}
{"id":"S17","category":"summarise","source":"articles/claude-vs-chatgpt-for-writing-2026.html","prompt":"Summarise the following passage in three short bullet points.\n\nMost comparisons of GPT -4 and later ChatGPT models versus Anthropic 's Claude for writing tasks settle on the same undifferentiated line: Claude is more literary, ChatGPT is more versatile. True, and useless, because nobody writes \"in general.\" You write a specific thing: a blog post, a novel chapter, a cold email, a brand-voice rewrite. So we did the work the consensus skips and scored both models task by task across nine real writing jobs, computed a per-use-case winner, then added the layer every writing-comparison page omits: the cheapest tier that actually unlocks the model that wins your job. A novelist and a marketer leave this page with different answers, and both know what to pay. Before you wade in: if you want to skip the prose and just plug the tasks into a side-by-side cost view, our AI tool pricing tracker keeps the live tier figures for both vendors in one place. Everything below is built on the same numbers."}
{"id":"S18","category":"summarise","source":"articles/best-ai-for-small-business.html","prompt":"Summarise the following passage in three short bullet points.\n\nRunning a small business in 2026 without AI is like running one in 2010 without a website, technically possible, but you're leaving an enormous competitive advantage on the table. The difference from the hype: you don't need a dozen AI subscriptions. You need the right AI tool for each part of your business, and most of the best ones are either free or under $20/month. This guide covers every business function, marketing, customer service, bookkeeping, hiring, legal, sales, and operations, with specific tool recommendations at every price point. For most small businesses, AI's biggest impact is in marketing, specifically, producing the content you know you should be creating but never have time for. Blog posts, social media, email newsletters, ad copy, product descriptions, all of this can be drafted by AI and refined by you in a fraction of the time."}
{"id":"S19","category":"summarise","source":"articles/how-to-read-an-ai-pricing-page-2026.html","prompt":"Summarise the following passage in three short bullet points.\n\nBefore any comparison, identify the unit. In one week we recorded ten: credits, AI credits, priority credits, Fast Tokens, characters, minutes of audio, minutes of video, videos per month, orders per month, recruiter seats, company headcount and annual revenue band. Two figures are only comparable if they price the same noun. Leonardo's is the one that proves it deliberate: its Token Bank is 25,500 on a plan granting 8,500 a month, 75,000 on 25,000, and 180,000 on 60,000. Three tiers, three different numbers, every one exactly three months of that tier's own allowance. All three vendors condition it on staying subscribed, and none extends it to a free plan."}
{"id":"S20","category":"summarise","source":"articles/deepseek-api-pricing-peak-hours-search-fee-2026.html","prompt":"Summarise the following passage in three short bullet points.\n\nDeepSeek's pricing page, read on 26 October 2026, says server-side search and fetch calls made through DeepSeek Harness, Claude Code or Codex are free today and will be billed at $0.75 per 1,000 successful requests from 00:00 Beijing time on 1 November 2026. The same page says off-peak rates are half the peak rates, and that two older model names are retired. DeepSeek's pricing page, read on 26 October 2026, says per-request billing starts at 00:00 Beijing time on 1 November 2026, at $0.75 per 1,000 successful requests for Search and $0.75 for Fetch. It says server-side calls made through DeepSeek Harness, Claude Code or Codex are currently free."}
{"id":"S21","category":"summarise","source":"articles/stable-diffusion-alternatives.html","prompt":"Summarise the following passage in three short bullet points.\n\nStable Diffusion is free, open, and endlessly customizable, but running it well means a capable NVIDIA GPU, Python environments, model downloads, and ongoing maintenance . If you would rather make images than manage a rig, these six cloud alternatives skip the hardware entirely, from the SD team's own next model to the easiest natural-language tools. Here is which one fits your work. Stable Diffusion's strengths are real: it is free, open-source, self-hostable, and infinitely tweakable with custom models, LoRAs, and extensions. The cost is not money, it is hardware and time . Running SD productively in 2026 needs a modern NVIDIA GPU with real VRAM, about 8GB minimum for SD 1.5 workflows and 12-16GB or more for SDXL and newer models at full quality, plus Python environments, model downloads, and ongoing troubleshooting as extensions and models update."}
{"id":"S22","category":"summarise","source":"articles/prompt-templates-and-variables.html","prompt":"Summarise the following passage in three short bullet points.\n\nA prompt template is the difference between a tool you use once and a tool you operate reliably. Write a bare instruction from scratch and you rewrite it every time, with no guarantee the next version behaves like the last. Build a template with named variable slots , a fixed output contract, and a self-critique loop baked in, and you get a durable asset that a team can share, version-pin, and improve without guesswork. This guide teaches the Variable-Slot Checklist (the centerpiece named asset), the four-layer anatomy, and the RAILS framework. It is part of our complete prompt engineering guide . A prompt template is a reusable instruction skeleton in which the parts that change across runs are marked as named variable slots and the parts that define the purpose, format, and quality bar are fixed. When you swap in concrete values for the slots, you get a fully specified, runnable prompt without rewriting the structure from scratch."}
{"id":"S23","category":"summarise","source":"articles/best-ai-tools-for-podcasters.html","prompt":"Summarise the following passage in three short bullet points.\n\nThe podcast tooling space in 2026 split into four distinct layers: recording, editing, post-production artifacts, and audio engineering . Almost no single tool wins all four, and the marketing copy on every vendor's site implies otherwise. We recorded and shipped 4 episodes of a 2-host test podcast (about 45 minutes each) through every tool below, ran each through its intended workflow, and timed the end-to-end cost-per-episode. This is what actually works, what the catches are, and the stack we would build at $0, $35, $60, $120, and $200+ per month. We recorded four episodes of a 2-host conversational test podcast, each running roughly 45 minutes, recorded remotely from two locations with mid-tier USB condensers (Shure MV7) plus video. Each episode then ran through one tool's full intended workflow end to end. We measured five things:"}
{"id":"S24","category":"summarise","source":"articles/is-claude-fable-5-worth-it.html","prompt":"Summarise the following passage in three short bullet points.\n\nAnthropic released Claude Fable 5 on June 9, 2026 and called it its most capable widely available model. The benchmarks back the headline: Fable 5 tops the charts on the hardest agentic coding. But it costs $10 per million input tokens and $50 per million output , which is exactly double Claude Opus 4.8 . This is a skeptic's review. After running both on real work, our read is blunt: for the vast majority of usage, Fable 5 is not worth twice the price, Opus 4.8 produces results close enough that you will not feel the gap, and the smart money runs a cheaper multi-model stack. Here is the math, the charts, and the exact setup that gets you Fable-tier output for far fewer tokens. Fable 5 is the public, safeguarded release of Anthropic's frontier line, shipped alongside a more restricted sibling, Mythos 5. It carries a 1 million token context window, up to 128,000 output tokens per response, adaptive thinking that is always on, and a January 2026 knowledge cutoff. It is available on the Claude API, on claude.ai across Pro, Max, Team, and Enterprise, in Claude Code and the desktop app, and through AWS, Google Cloud, Microsoft Foundry, and GitHub Copilot. In other words, it is everywhere Opus already was."}
{"id":"S25","category":"summarise","source":"articles/claude-pro-annual-vs-monthly-billing-2026.html","prompt":"Summarise the following passage in three short bullet points.\n\nMost pages bury this decision in a single table cell: \" $17/mo, billed annually ,\" and move on. That cell hides the actual question. Anthropic charges $200 upfront for annual Claude Pro versus $20 rolling for monthly, so this is not really a price comparison, it is a bet on your own future usage. You save $40 a year if you stay, and you forfeit the unused months if you do not. Below we run the break-even no commodity pricing table runs: the exact crossover month where annual starts winning, and the rule for whether you should take the bet at all. The only thing annual billing changes is the billing cadence and the price, not the product. On the official Claude pricing page , Pro is listed at $20/month billed monthly verified 2026-06-10 . Switch the toggle to annual and the same plan reads $17/month verified 2026-06-10 , charged as a single $200 payment up front verified 2026-06-10 for the year."}
{"id":"S26","category":"summarise","source":"articles/best-free-ai-art-generators-2026.html","prompt":"Summarise the following passage in three short bullet points.\n\nIf you want art , not just a stock-photo replacement, the ranking changes. The best free AI art generators are the ones with the deepest style range: anime, fantasy, painterly, and concept-art looks you can actually direct. We tested each with style-heavy prompts and ranked them by free output, aesthetic control, and how close they get to a deliberate artistic look for $0. (For a utility-first list, see our best free AI image generators guide.) For art, two things matter more than they do for stock-style images: style depth and style control . Style depth is how many distinct artistic looks a tool can produce well, anime versus oil-painting versus pixel art versus concept sketch. Style control is how precisely you can steer it toward the exact look in your head, through style models, reference images, fine-tunes, or LoRAs."}
{"id":"S27","category":"summarise","source":"articles/claude-code-vs-cursor-vs-copilot.html","prompt":"Summarise the following passage in three short bullet points.\n\nThree AI coding agents now dominate professional developer workflows. Claude Code from Anthropic runs as a CLI agent in your terminal and rates as the most loved developer tool of 2026 in the Stack Overflow survey. Cursor is a full AI-native IDE that replaces VS Code entirely. GitHub Copilot is the plugin that lives inside the editor you already use. Same mission, three completely different surfaces. We ran the same agentic task through each of them. Here is the side-by-side, the pricing, and a clear pick by workflow. Read of the result. All three completed the task with passing tests. Claude Code was the fastest and most autonomous: one approval covered the entire plan. Cursor was second with plan-first execution and a small bonus prompt to confirm running the test. Copilot was the slowest but the safest, with five approvals giving the human a checkpoint per file. The right answer depends entirely on whether you want autonomy or control."}
{"id":"S28","category":"summarise","source":"articles/ai-dubbing-tools-for-ad-creatives-2026.html","prompt":"Summarise the following passage in three short bullet points.\n\nSeven AI dubbing tools compared on the thing an ad team buys: a localised 15 to 60 second spot. Each is priced per finished minute from its own published meter, with voice cloning, lip-sync and commercial rights shown for each. Every price was read on the vendor’s own page on 26 October 2026. Read from each tool’s own published meter, the lowest price of one finished minute of AI dubbing runs from $0.16 (Speechify Studio Creator) to $6.83 (AdsTurbo Pro, yearly), a 44-fold spread; six of seven vendors print a meter, HeyGen does not, and where lip-sync is priced it multiplies the minute (Synthesia: 2x credits)."}
{"id":"S29","category":"summarise","source":"articles/best-ai-email-assistants-2026.html","prompt":"Summarise the following passage in three short bullet points.\n\nAI email assistants in 2026 range from passive drafting helpers bolted onto Gmail to semi-autonomous agents that read, sort, draft, and conditionally send on your behalf. The gap between those two ends of the spectrum is enormous, and so is the price gap. The short answer: Superhuman for power users who want speed and draft quality; Fyxer AI for voice-matched autonomous drafts; Shortwave if you want a clean Gmail replacement with AI baked in; Copilot or Gemini if you are already inside Microsoft 365 or Google Workspace; SaneBox for triage-only; Jace if you want the most autonomous option available today. Use our AI stack optimizer to find the right fit in 30 seconds based on your email volume and client. A basic email filter moves messages into folders using rules you wrote. An AI email assistant does something meaningfully different: it reads the content of messages, understands context and intent, and either suggests or takes action based on that understanding. The 2026 generation of these tools splits into three tiers."}
{"id":"S30","category":"summarise","source":"articles/best-chatgpt-alternatives-2026.html","prompt":"Summarise the following passage in three short bullet points.\n\nWith ChatGPT drawing more complaints and cancellations in 2026, the good news is the alternatives are genuinely excellent, and several beat it at specific jobs. Claude leads for writing and coding, Perplexity for sourced research, and Gemini for a free, Google-integrated everyday assistant. There is no single \"best,\" only the best switch for what you do most. We tested the field. Here are the strongest ChatGPT alternatives in 2026, ranked, plus one you should approach with caution. Switch by task, not by reputation. If you write or code, Claude is the strongest move and the most common one. If you do research, Perplexity's cited answers beat a plain chatbot. If you want the best free experience or live in Google's apps, Gemini wins. If your work is in Microsoft 365, Copilot is the natural fit. Many people run two or three, Claude for analysis, Perplexity for sourced research, and a free option for quick tasks, which maximizes free tiers without being locked to one platform. Below, the best alternative for each job."}
{"id":"W01","category":"long-form","prompt":"Explain why the sky is blue for a curious teenager. Aim for about 400 words and use plain language."}
{"id":"W02","category":"long-form","prompt":"Explain how vaccines train the immune system for a busy adult with no background in the subject. Aim for about 400 words and use plain language."}
{"id":"W03","category":"long-form","prompt":"Explain what compound interest is for a first-year college student. Aim for about 400 words and use plain language."}
{"id":"W04","category":"long-form","prompt":"Explain how a bill becomes law in the United States for a curious teenager. Aim for about 400 words and use plain language."}
{"id":"W05","category":"long-form","prompt":"Explain how noise cancelling headphones work for a busy adult with no background in the subject. Aim for about 400 words and use plain language."}
{"id":"W06","category":"long-form","prompt":"Explain what inflation is and what causes it for a first-year college student. Aim for about 400 words and use plain language."}
{"id":"W07","category":"long-form","prompt":"Explain how the water cycle works for a curious teenager. Aim for about 400 words and use plain language."}
{"id":"W08","category":"long-form","prompt":"Explain what a mortgage amortisation schedule shows for a busy adult with no background in the subject. Aim for about 400 words and use plain language."}
{"id":"W09","category":"long-form","prompt":"Explain why the Roman Empire declined for a first-year college student. Aim for about 400 words and use plain language."}
{"id":"W10","category":"long-form","prompt":"Explain how solar panels produce electricity for a curious teenager. Aim for about 400 words and use plain language."}
{"id":"W11","category":"long-form","prompt":"Explain the difference between a Roth IRA and a traditional IRA for a busy adult with no background in the subject. Aim for about 400 words and use plain language."}
{"id":"W12","category":"long-form","prompt":"Explain how GPS determines location for a first-year college student. Aim for about 400 words and use plain language."}
{"id":"W13","category":"long-form","prompt":"Explain what causes the seasons for a curious teenager. Aim for about 400 words and use plain language."}
{"id":"W14","category":"long-form","prompt":"Explain how antibiotics work and why resistance develops for a busy adult with no background in the subject. Aim for about 400 words and use plain language."}
{"id":"W15","category":"long-form","prompt":"Explain how the stock market sets prices for a first-year college student. Aim for about 400 words and use plain language."}
{"id":"W16","category":"long-form","prompt":"Explain what machine learning is for a curious teenager. Aim for about 400 words and use plain language."}
{"id":"W17","category":"long-form","prompt":"Explain how a refrigerator keeps food cold for a busy adult with no background in the subject. Aim for about 400 words and use plain language."}
{"id":"W18","category":"long-form","prompt":"Explain what the Industrial Revolution changed for a first-year college student. Aim for about 400 words and use plain language."}
{"id":"W19","category":"long-form","prompt":"Explain how the internet routes data between computers for a curious teenager. Aim for about 400 words and use plain language."}
{"id":"W20","category":"long-form","prompt":"Explain how credit scores are calculated for a busy adult with no background in the subject. Aim for about 400 words and use plain language."}
{"id":"W21","category":"long-form","prompt":"Explain what causes earthquakes for a first-year college student. Aim for about 400 words and use plain language."}
{"id":"W22","category":"long-form","prompt":"Explain how photosynthesis works for a curious teenager. Aim for about 400 words and use plain language."}
{"id":"W23","category":"long-form","prompt":"Explain why bridges are built with arches or trusses for a busy adult with no background in the subject. Aim for about 400 words and use plain language."}
{"id":"W24","category":"long-form","prompt":"Explain how a bicycle stays upright for a first-year college student. Aim for about 400 words and use plain language."}
{"id":"W25","category":"long-form","prompt":"Explain what a blockchain is for a curious teenager. Aim for about 400 words and use plain language."}
{"id":"W26","category":"long-form","prompt":"Explain how the human heart pumps blood for a busy adult with no background in the subject. Aim for about 400 words and use plain language."}
{"id":"W27","category":"long-form","prompt":"Explain what causes ocean tides for a first-year college student. Aim for about 400 words and use plain language."}
{"id":"W28","category":"long-form","prompt":"Explain how a jet engine works for a curious teenager. Aim for about 400 words and use plain language."}
{"id":"W29","category":"long-form","prompt":"Explain what supply and demand mean for a busy adult with no background in the subject. Aim for about 400 words and use plain language."}
{"id":"W30","category":"long-form","prompt":"Explain how memory works in the human brain for a first-year college student. Aim for about 400 words and use plain language."}
{"id":"R01","category":"reasoning","prompt":"A shop sells notebooks for $7.75 each. Maria buys 15 of them and pays with a $130 bill. How much change does she get? Show your steps."}
{"id":"R02","category":"reasoning","prompt":"A train travels 235 km in 9 hours and then 170 km in 10 hours. What is its average speed for the whole trip? Show your steps."}
{"id":"R03","category":"reasoning","prompt":"A recipe needs 6 cups of flour for 14 muffins. How many cups are needed for 42 muffins, and how many cups for 19 muffins? Show your steps."}
{"id":"R04","category":"reasoning","prompt":"If 2 workers can build a wall in 21 days, how many days would 4 workers need, assuming they all work at the same rate? Show your steps."}
{"id":"R05","category":"reasoning","prompt":"A shop sells notebooks for $11.00 each. Maria buys 12 of them and pays with a $150 bill. How much change does she get? Show your steps."}
{"id":"R06","category":"reasoning","prompt":"A train travels 205 km in 8 hours and then 100 km in 9 hours. What is its average speed for the whole trip? Show your steps."}
{"id":"R07","category":"reasoning","prompt":"A recipe needs 4 cups of flour for 21 muffins. How many cups are needed for 63 muffins, and how many cups for 26 muffins? Show your steps."}
{"id":"R08","category":"reasoning","prompt":"If 2 workers can build a wall in 79 days, how many days would 4 workers need, assuming they all work at the same rate? Show your steps."}
{"id":"R09","category":"reasoning","prompt":"A shop sells notebooks for $6.50 each. Maria buys 13 of them and pays with a $100 bill. How much change does she get? Show your steps."}
{"id":"R10","category":"reasoning","prompt":"A train travels 385 km in 8 hours and then 110 km in 9 hours. What is its average speed for the whole trip? Show your steps."}
{"id":"R11","category":"reasoning","prompt":"A recipe needs 6 cups of flour for 14 muffins. How many cups are needed for 42 muffins, and how many cups for 19 muffins? Show your steps."}
{"id":"R12","category":"reasoning","prompt":"If 9 workers can build a wall in 26 days, how many days would 11 workers need, assuming they all work at the same rate? Show your steps."}
{"id":"R13","category":"reasoning","prompt":"A shop sells notebooks for $5.25 each. Maria buys 15 of them and pays with a $90 bill. How much change does she get? Show your steps."}
{"id":"R14","category":"reasoning","prompt":"A train travels 235 km in 5 hours and then 200 km in 6 hours. What is its average speed for the whole trip? Show your steps."}
{"id":"R15","category":"reasoning","prompt":"A recipe needs 7 cups of flour for 5 muffins. How many cups are needed for 15 muffins, and how many cups for 10 muffins? Show your steps."}
{"id":"R16","category":"reasoning","prompt":"If 3 workers can build a wall in 62 days, how many days would 5 workers need, assuming they all work at the same rate? Show your steps."}
{"id":"R17","category":"reasoning","prompt":"A shop sells notebooks for $5.50 each. Maria buys 20 of them and pays with a $120 bill. How much change does she get? Show your steps."}
{"id":"R18","category":"reasoning","prompt":"A train travels 120 km in 4 hours and then 50 km in 5 hours. What is its average speed for the whole trip? Show your steps."}
{"id":"R19","category":"reasoning","prompt":"A recipe needs 5 cups of flour for 10 muffins. How many cups are needed for 30 muffins, and how many cups for 15 muffins? Show your steps."}
{"id":"R20","category":"reasoning","prompt":"If 6 workers can build a wall in 40 days, how many days would 8 workers need, assuming they all work at the same rate? Show your steps."}
{"id":"R21","category":"reasoning","prompt":"A shop sells notebooks for $18.00 each. Maria buys 7 of them and pays with a $140 bill. How much change does she get? Show your steps."}
{"id":"R22","category":"reasoning","prompt":"A train travels 250 km in 5 hours and then 180 km in 6 hours. What is its average speed for the whole trip? Show your steps."}
{"id":"R23","category":"reasoning","prompt":"A recipe needs 9 cups of flour for 13 muffins. How many cups are needed for 39 muffins, and how many cups for 18 muffins? Show your steps."}
{"id":"R24","category":"reasoning","prompt":"If 9 workers can build a wall in 71 days, how many days would 11 workers need, assuming they all work at the same rate? Show your steps."}
{"id":"R25","category":"reasoning","prompt":"A shop sells notebooks for $15.00 each. Maria buys 12 of them and pays with a $190 bill. How much change does she get? Show your steps."}
{"id":"R26","category":"reasoning","prompt":"A train travels 390 km in 3 hours and then 140 km in 4 hours. What is its average speed for the whole trip? Show your steps."}
{"id":"R27","category":"reasoning","prompt":"A recipe needs 5 cups of flour for 10 muffins. How many cups are needed for 30 muffins, and how many cups for 15 muffins? Show your steps."}
{"id":"R28","category":"reasoning","prompt":"If 5 workers can build a wall in 87 days, how many days would 7 workers need, assuming they all work at the same rate? Show your steps."}
{"id":"R29","category":"reasoning","prompt":"A shop sells notebooks for $18.25 each. Maria buys 10 of them and pays with a $200 bill. How much change does she get? Show your steps."}
{"id":"R30","category":"reasoning","prompt":"A train travels 155 km in 4 hours and then 140 km in 5 hours. What is its average speed for the whole trip? Show your steps."}