Do More Newsletter

Sponsored by

Keep up to date on the latest products, workflows, apps and models so that you can excel at your work. Curated by Duet.

Stay ahead with the most recent breakthroughs—here’s what’s new and making waves in AI-powered productivity:

Perplexity AI

Perplexity AI offers an intelligent alternative to traditional web search by delivering clear, cited answers to complex queries in real time. Designed for researchers, consumers, and business owners, it aggregates information from trusted web sources to create structured summaries. The latest platform update enhances deep research capabilities, allowing users to analyze technical reports, compare product specifications, and generate data spreadsheets in seconds.

Suno AI

Suno AI brings music creation to everyday consumers and content creators through intuitive text prompts. Users can describe a genre, mood, or lyric idea, and the platform generates full fidelity audio tracks complete with vocals and instrumentation. The newest platform enhancements give users precise control over song structure, enabling creators to extend tracks, adjust tempo, and produce custom background audio for social media, podcasts, or personal projects.

Typefully

Typefully is a social media growth platform tailored for creators and small business marketers who want to craft and schedule high performing content across major channels. Utilizing conversational AI models trained on viral writing techniques, the tool helps draft post threads, rewrite drafts, and analyze audience engagement analytics. The recent update introduces automated content scheduling driven by audience activity trends, helping posts reach maximum visibility without manual oversight.

Descript

Descript simplifies video and podcast editing by converting recorded media into readable text documents. Creators can edit video or audio recordings simply by deleting or modifying words in the written transcript, eliminating the need for complex timeline editing. Recent media coverage focuses on Descript's upgraded AI voice synthesis and background noise reduction, which allows users to fix misspoken words seamlessly using generated voice clones and clean up poor microphone audio with a single click.

TubeBuddy

TubeBuddy provides creators and small business marketers with an intelligent platform designed to streamline YouTube channel management and video strategy. By analyzing channel performance data, viewer comment trends, and audience behavior, the tool generates tailored content recommendations and growth strategy insights. The latest AI feature delivers data backed video recommendations based on proprietary channel signals, helping creators craft engaging video concepts, optimize titles, and boost subscriber retention.

Perplexity AI Elevates Product Research and Discovery

Perplexity AI has transformed from a simple search alternative into an essential knowledge engine for everyday consumers and small business teams. By combining live web indexing with conversational reasoning, the platform delivers direct, cited answers to complex questions, bypassing traditional search engine spam and cluttered result pages.

The latest update introduces Pro Search Assistant, a streamlined shopping and discovery capability that actively researches products, compares price histories, and summarizes user reviews across major online platforms. Rather than opening dozens of browser tabs, users can prompt the system with detailed specifications to instantly receive a curated buying guide tailored to their budget and performance preferences.

This new feature provides immense time savings for those looking to make informed purchasing decisions without wasting hours filtering through sponsored links. Small businesses can quickly evaluate hardware, software subscriptions, or office gear, while everyday users get unbiased product recommendations grounded in authentic customer feedback from reputable platforms.

Crucially, the capability allows users to set custom preference filters, such as prioritizing eco friendly manufacturing, local availability, or specific warranty terms. The system also generates interactive comparison tables that export directly into spreadsheets, making team hardware procurement or personal budgeting seamless and effortless.

Every Chatbot Explained: ChatGPT vs. Claude vs. Gemini vs. Grok

“Which AI should I actually use?” might be the most common question in tech right now — and it’s weirdly hard to answer, because the honest answer keeps changing, making this question worth revisiting maybe multiple times per year.

Take just this past week. Google’s next flagship model slipped again — months late and still not out. European regulators ordered Google to let rival AI assistants onto Android phones. Anthropic shipped a brand-new flagship model, capping a month of churn over which model paying subscribers even get. And a Chinese lab most people have never heard of topped a major coding leaderboard, ahead of the big names — then said it would give the model away for anyone to run.

That’s one week. It goes like this most weeks.

So there’s no single answer that stays true. But there’s a way to think about the question that survives the churn: stop asking “which one is best” — a question with no stable answer — and start asking “which one is best at the thing I do most.” Answer that, and the choice mostly makes itself.

Here are the big four — the names everyone means when they say “the chatbots”: ChatGPT, Claude, Gemini, and Grok. Below is what each is genuinely good at, the smartest model each one is running right now, and when to reach for it.

One thing first, because it’s the most freeing fact in this whole piece: the free version of every one of these is genuinely good now. You do not have to pay anyone twenty dollars to find out what you like. Sign up for two or three for free, use them for a week, and let the paid decision come to you.

ChatGPT — The Safe Default

If you sign up for exactly one and never think about it again, this is the one. ChatGPT is what most people picture when they hear “AI,” and there’s a reason more than 900 million people use it every week — its app passed a billion monthly users this spring, the fastest any app has ever hit that mark.

The vibe: the capable generalist. Quick, confident, does a little of everything well.

Smartest model right now: GPT-5.6 Sol, the top “thinking” model in the family OpenAI shipped earlier this month. (Sol is the flagship; its faster, cheaper siblings are named Terra and Luna, and most everyday chatting runs on a quick default model — you rarely have to pick.)

What it’s great at. The broadest range of anything on this list. It writes, it explains, it makes images (just describe what you want and it draws it), it reads documents you upload, it crunches a spreadsheet, it searches the web when your question needs current facts. Its top models are strong at the kind of slow, step-by-step reasoning that trips lesser tools up. And its voice mode got a real upgrade this month — you can now talk over it, interrupt it, and have it actually stop and listen, the way a person does. It finally feels less like a walkie-talkie and more like a phone call.

Where it falls short. Being good at everything means it’s occasionally the best at nothing. Serious writers still find its prose a touch generic next to Claude. And OpenAI ships so fast that things move under your feet — the current models are only a few weeks old, and the lineup shifts constantly.

Reach for it when: you want one dependable assistant for a bit of everything, you want to make images, or you want to talk to your AI out loud. Free tier is excellent; Plus is $20 a month (there’s also a lighter $8 “Go” tier) if you want the smartest models and the fancier tools.

Claude — The One That Writes Like a Human

Claude, made by Anthropic, has quietly become a favorite of writers, editors, and just about anyone who works with words for a living. Here’s why.

The vibe: the thoughtful one. Reads the room, hates fluff, tells you when the answer is honestly “it depends.”

Smartest model right now: Claude Opus 5, which landed just days ago. Anthropic says it comes close to the raw intelligence of its heavyweight Fable 5 model at half the cost — and, more to the point for the rest of us, it’s the strongest model on the regular $20 Pro plan. The free tier runs the very capable Claude Sonnet 5.

What it’s great at. Two things, better than most. First, writing — if you want an email, a toast, a tricky note to your landlord, or anything that has to sound like a human wrote it, Claude’s output tends to need the least cleanup of any tool here. Second, anything long or nuanced. You can hand it an entire contract, a dense report, or a book’s worth of text and it’ll hold the whole thing in its head and answer questions about page 200. For decisions where a too-confident answer would actually steer you wrong — money, health, “should I take this job” — it walks you through the trade-offs instead of barking a verdict.

Where it falls short. It won’t draw you a picture — Claude deliberately doesn’t do image generation, betting its chips instead on writing, reasoning, and building things (it can whip up a working mini webpage or a slide mockup, just not a photo of a golden retriever). And it’s a beat slower to hand you a clean one-word recommendation, because it genuinely wants to think first.

Reach for it when: you’re writing something that matters, wrestling with a long document, or thinking through a decision with real nuance. Free tier runs its capable everyday model; Pro is about $20 a month. (A small sign of how fast this world moves: Anthropic spent a solid month of shifting deadlines sorting out how subscribers get its heavyweight Fable 5 model — and then dropped an even newer flagship, Opus 5, on top of it. That happened while this article was being written. Blink and the details change.)

Gemini — The One Already in Your Stuff

This is the one a lot of people should probably actually use, because their whole life already lives in Gmail and Google Docs — and Gemini is Google’s AI, built to plug right into all of it.

The vibe: the helpful colleague who — once you let it — has access to your files.

Smartest model right now: Gemini 3.1 Pro — Google’s best released model. Confusingly, a more powerful “Gemini 3.5 Pro” has been promised for months and keeps slipping (more on that in a second).

What it’s great at. Living where you already live. With your permission, it can connect to Gmail, Drive, and Calendar — so you can ask it to find the resume in your Drive, summarize the unread email from your boss, or pull the dates out of a thread and drop them on your calendar, and it’ll actually go do it. It’s also the best of the bunch at video and big media — you can feed it an hour-long YouTube video and get the gist in seconds — and it makes images and short video clips too. And on many Android phones, Gemini is already built in, or a free download away.

Where it falls short. Google is having a genuinely rough stretch. That more powerful model just above was promised months ago and still isn’t out — it’s been pushed back again and again while Google reworks it, so the company is in a holding pattern while competitors ship. Its writing tends to run drier than Claude’s. And the interface can feel busy. One bit of good news, at least for Android users in Europe: regulators there just ordered Google to give rival assistants real access to Android — voice activation, actions inside apps — which should make it much easier to use the assistant you actually want.

Reach for it when: your work and life run on Google, or you deal with a lot of video. Free tier is real; paid plans start around $5 a month for a light tier and $20 for the full “AI Pro” experience.

Grok — The Wildcard

And then there’s Grok, Elon Musk’s AI — these days built by SpaceXAI, after SpaceX absorbed xAI earlier this year — and still closely tied to X (the site formerly known as Twitter). It’s the odd one out here, on purpose.

The vibe: the one that talks back, cracks jokes, and skips the lecture.

Smartest model right now: Grok 4.5, which launched earlier this month and which the company pitches as its smartest yet — independent testers put it near the top of the pack.

What it’s great at. Real-time, right-now information. Because it’s plugged straight into the X firehose, Grok often knows about breaking news — a trade, a scandal, a game result — while the other tools are still catching up. It has a looser, funnier, more opinionated personality that some people genuinely prefer to the buttoned-up competition, it’ll generate images with fewer guardrails than most, and xAI has leaned hard into chatty “AI companion” characters.

Where it falls short. You can use it free at grok.com or in its own app now — no X account required — though the best version still sits behind a paid plan. That same loose personality that’s fun can also veer into edgy or just plain wrong, and on ordinary factual questions it has historically trailed ChatGPT and Claude. It’s a spice, not the main dish.

Reach for it when: you want up-to-the-second social and news chatter, or you just want an AI with an attitude. There’s a free tier; the SuperGrok upgrade runs about $30 a month — a bit more than the others’ $20 plans.

A Word on the Newcomers

You’ll start hearing two more names — DeepSeek and Kimi, two Chinese AI outfits — so it’s worth knowing what they are before the headlines rattle you. Their whole strategy is to be radically cheap or outright free; Kimi’s latest just topped a major coding leaderboard, ahead of models from OpenAI and Anthropic, and the company says it’ll release the model free for anyone to run.

You can try them today — Kimi runs a free chatbot on its website, just like the others — but for most people they’re not a daily driver yet. Where they really matter to you is subtler: their existence is why the tools above keep getting cheaper and better. Competition is doing exactly what competition is supposed to do. You benefit whether or not you ever touch them.

(And if your main goal is searching the web and getting answers with sources, there’s another tool worth a look — Perplexity.)

The Cheat Sheet

If you remember nothing else, remember this:

• Want one tool for a bit of everything (or to make images, or to talk out loud)? ChatGPT.

• Writing something that matters, or thinking through a big decision? Claude.

• Live inside Gmail, Docs, and Drive — or work with a lot of video? Gemini.

• Want news the second it breaks, with some attitude? Grok.

That’s it. Tape it to the fridge.

A Few Honest Caveats

First: they all still make things up. Every single one, ChatGPT and Claude and Gemini and Grok alike, will occasionally state something wrong with the exact same friendly confidence it uses when it’s right. They’ve improved a lot, but they’re far from perfect. For anything that touches your health, your money, or a date you can’t miss, treat the answer as a very smart first draft — and verify before you act on it.

Second: don’t over-subscribe. A common mistake is paying $20 a month to three different companies out of a vague fear of missing out. You almost certainly don’t need that. Pick your one main tool based on the cheat sheet above, keep a free account or two on the side for the occasional second opinion, and put the other forty dollars back in your pocket.

Third: mind what you share. Before you connect an assistant to your email or files — or paste in a contract, a medical result, or anything sensitive — check the app’s privacy settings, and if it’s for work, check your company’s rules. These tools are convenient right up until they’re holding something you didn’t mean to hand over.

The Bottom Line

The real answer to “which one should I use” isn’t a name — it’s a habit. Notice what you actually do with an AI most days. Writing? Reach for Claude. Living in Google? Gemini. A little of everything? ChatGPT. Glued to the news? Grok, and good luck.

The models will keep leapfrogging each other — there’ll be a new “smartest one” by next month, and the rankings above will shift. That’s fine. You’re not marrying one. You’re just picking the right tool for the job in front of you, the same way you already reach past the butter knife for the scissors when the job calls for scissors.

Pick the one that fits what you do most. Keep a free account or two around for second opinions. And when the lineup reshuffles again next week — it will — you won’t have to care, because your answer won’t be a brand name. It’ll be a question: what am I actually trying to do?

Partner Spotlight: AutoDoc by Duet Display

Meeting documentation just became far more effective and private with AutoDoc by Duet Display. AutoDoc is a completely free application that runs locally on your device, operating smoothly even when you are disconnected from the Internet. Because all processing stays on your machine, your private information never leaves your computer, providing complete privacy for sensitive business discussions or personal notes.

AutoDoc records both audio and video during your sessions, automatically syncing your written notes directly to the video timeline. This unique feature means you can read what happened during a meeting while simultaneously viewing the exact visual presentation delivered at that precise moment. AutoDoc is an open source project, but a convenient installer download is available for users who prefer a quick, straightforward setup without touching code.

Download AutoDoc for free at: getAutodoc.com

The AI Agent You Can Trust

The best assistants don't multitask their attention across a hundred tools. Neither does Catch. It's an AI agent that focuses on one thing — the admin work you'd rather not touch — and does it exceptionally well.

Scheduling, flights, restaurants, follow-ups, vendors, clients. You hand it over; Catch handles the back-and-forth and comes back with it done.

No context-switching. No dropped balls. Just your admin, quietly cleared — so your focus stays on the work only you can do.

Meet the agent built for admin, and it'll be ready to work before your next meeting.

Get started at catchagent.ai — and give your attention back to what matters.

Start now for free

Stay productive, stay curious—see you next week with more AI breakthroughs!