Editor’s Note: Plain Manual Weekly #7, our Saturday roundup. Every week: everything we published, plus the biggest AI and software stories, translated into plain English. Evidence label: news roundup. The four mini-stories below synthesize this week’s reporting, with a source line under each story. They contain no hands-on testing. The articles recapped below went through our daily pipeline.
Grab a coffee. Here’s your week. The smartest AI model changed hands again, ChatGPT opened its doors to businesses and then gave its agents a wallet, and Google brought its newest Gemini to everyone. Meanwhile, we published fourteen guides, including our 100th post.
What we opened this week
- Monday’s SEO loop got a signed-in Search Console read at last: 2,684 impressions and 7 clicks over 28 days, and it added TPM Scorecards to Fathom vs Otter.ai (7.4 to 7.2) and FreshBooks vs Wave (7.4 to 6.4) after rechecking both pairs’ prices on the makers’ own pages, with no price changes found.
- Alternatives to Squarespace published on Wednesday, and its cohort welcome page, New Here From Squarespace? Start Here, went live with it as the first use of that template.
- Thursday brought two milestones in one day: the 50 best lines from our first 100 posts, and the weekly X thread, on Bitwarden vs LastPass.
- On Friday the new Time-Zone Meeting Planner tool went live, then got a same-day rework after a phone check caught overlapping fields and a missed Los Angeles to Sydney next-day window. The rework passed all 20 checks.
- Reader Poll #1 closed this week with 0 responses across October 20 to November 10. We removed its callout from the Subscription Cost Calculator page and said so plainly. Thin audience, honest count.
- On Saturday, ClickUp’s own pricing FAQ gave us a finding we did not expect: on the free plan, guests can edit everything, and that permission cannot be turned down.
Plain in 2 Minutes: 3 verdicts, 2 price moves, 1 question
3 verdicts:
- Trello Free is the pick for most freelancers: unlimited cards, 10 boards, and room for clients without paying.
- For most people who just want privacy, a short list of VPNs earns its keep, and the rest of the market is noise. Our Best VPN guide names the documented picks.
- If Squarespace’s July price rise pushed you out, the leaving guide groups the alternatives by your reason for leaving, with a migration checklist at the end.
2 price moves:
- Claude Fable 5.2 launched at $8 per million input tokens, while Meta’s Muse Ember launched at $0.10 for repetitive agent work. The gap between planning and doing is now a price list.
- TickTick Premium now lists at $49.99 a year, not the $35.99 in older guides, including some of ours. Check the list price, not the discount screenshot.
1 question: Would you let an AI agent pay for something with your card if you approved every purchase first? Story 4 is that exact offer, now live for some people.
Number of the week: 65. That is Claude Fable 5.2’s score on the Artificial Analysis Intelligence Index, ahead of Gemini 4 Carbon at 60. A benchmark lead is real and narrow at the same time. Story 1 has the details.
This week’s news, in plain English
1. Claude Fable 5.2 took the top score, and Meta priced the grunt work at ten cents
Anthropic released Claude Fable 5.2 on Monday, November 9, alongside a restricted sibling, Claude Mythos 5.2, for vetted cybersecurity and life sciences organizations. Fable 5.2 scored 65 on the Artificial Analysis Intelligence Index, ahead of Google’s Gemini 4 Carbon at 60 and Anthropic’s own Claude Opus 5.5 at 59. Anthropic says it produces output up to 30 percent faster than Fable 5.1 and that its safety filters block about 55 percent fewer legitimate requests. The price is $8 per million input tokens and $40 per million output tokens. A token is the small chunk of text AI companies use to meter usage.
The same day, Meta released Muse Ember, its cheapest model yet, at $0.10 per million input tokens. Ember is built to do repetitive steps, such as tool calls, data extraction, and long-document reading, while a larger model plans. In plain English, the industry is splitting the job in two: one model to think, a much cheaper one to do.
Our take, through the Receipts Test: Receipts partial. The Work has receipts, an independent index score and named users such as Box, which used the model to review loan files. The Numbers are published prices, which is a real receipt. The Catch is that a benchmark lead tells you little about your own task, and 65 against 60 is a lead, not a landslide.
Source: PYMNTS, Anthropic’s Claude tops AI rankings (November 12, 2026)
2. Businesses can now claim their profile inside ChatGPT
OpenAI opened Businesses in ChatGPT in beta on Tuesday, November 10. A company can claim its profile in a new portal called Business Hub, verify that it owns or represents the business, correct the details ChatGPT shows, and add product feeds with prices and availability. The Hub also shows aggregated figures: how often ChatGPT mentioned the business and how many people clicked through to its website, over seven-day and 28-day views.
Eligible ecommerce and online service businesses in the United States can go one step further and publish a Business Agent, a branded assistant that answers customer questions inside ChatGPT using information the business provides. OpenAI is clear about the limit that matters: claiming a profile does not guarantee higher placement or more sales, and businesses never see individual customer conversations.
Our take: this is the business version of claiming a listing on a maps service, moved into the answer box. If customers ask ChatGPT about your kind of business, the details it repeats are now partly yours to correct. The beta label, the US-only agent step, and the no-guarantee line are all worth keeping in view.
Source: PPC Land, Businesses gain ChatGPT click and mention counts in OpenAI’s Business Hub (November 2026)
3. Gemini 4 Hydrogen brought Google’s newest model to everyone
Google introduced Gemini 4 Hydrogen on Thursday, November 12, describing it as cutting-edge intelligence for everyone, built for speed and efficiency in everyday coding and agent work. It rolls out across Google AI Studio, the Gemini API, Gemini for business, the Gemini app, Google Search, and Antigravity for paid Pro and Ultra subscribers. Google says Hydrogen surpasses its Argon model on the Artificial Analysis Intelligence Index at its default thinking level, 51 to 50, and reaches 54 at its highest setting.
The price is flat across its full 1 million-token context window: $1.50 per million input tokens and $7.50 per million output tokens. The context window is how much text the model can take in at once. Hydrogen offers four thinking levels, from low to a new extra-high tier, so a user pays for depth only when a task needs it.
Our take: for ordinary readers, the model name matters less than where it shows up. Hydrogen is the version of Gemini 4 that lands in Search and the Gemini app, which is where most people will meet it. Judge it on whether your everyday answers get more reliable, not on the one-point index gap.
Source: Google, Introducing Gemini 4 Hydrogen (November 12, 2026)
4. ChatGPT Wallet lets an agent pay, but only after you approve each purchase
OpenAI launched ChatGPT Wallet on Friday, November 13. It works with OpenAI’s dot agents and was built with American Express, Mastercard, and Visa. The wallet lets an agent complete a purchase without ever seeing your actual card details. You add a card once. Before each purchase, you review the items, the total, and the selected card, then approve with a security check such as Face ID or a fingerprint scan.
After approval, the payment network issues a payment token, a substitute card number, security code, and expiry date, and the charge lands on your selected card. OpenAI says existing card rewards and dispute protections are preserved. At launch the wallet is limited to ChatGPT Pro subscribers who are 18 or older in the United States with eligible US-issued cards, and it will not work on sites that block agents.
Our take, through the Receipts Test: Receipts shown for the design, with the rollout as the catch. The Work and the Numbers are concrete: per-purchase approval, tokenized payment, named card partners. The Catch is scope. Pro only, US only, eligible cards only, and a merchant can still refuse an agent at checkout. Approving every purchase is the right default, and it is the detail to watch if it ever changes.
Source: 9to5Mac, OpenAI launches ChatGPT Wallet (November 13, 2026)
This week’s articles
Fourteen new posts, November 8 to 14. Start with these three:
- Alternatives to Squarespace, After the July 2026 price increase, the picks grouped by why you are leaving, plus a migration checklist.
- The Best VPN for Normal People, Documented pricing and limits only, for readers who want privacy without a hobby.
- The Best Free Project Management Tools for Freelancers, Trello Free leads, ClickUp gives the most machinery at $0, and monday.com’s free tier is a tryout only.
Versus
- TickTick vs Microsoft To Do, Which free to-do app should you use, priced and compared on documented facts.
- WordPress vs Squarespace, Which one owns your website, and what leaving each one really involves.
Best For
- The Best AI Tools for Job Seekers, The field priced out, including the weekly-billing traps hiding in plain sight.
- 5 AI Tools Actually Worth Paying For, Five that earn a subscription, and the rest you can skip.
TIL: short notes
- TIL: Teal+ Costs $13 a Week, Which adds up to $676 a year.
- TIL: TickTick Premium Now Lists at $49.99 a Year, Not the $35.99 in older guides.
- TIL: Mullvad Has Never Had a Sale, And its price has not changed since 2009.
- TIL: Squarespace Only Exports in WordPress Format, And your store pages stay behind.
- TIL: Grammarly Pro Costs $360 a Year Paid Monthly, And $144 if you pay yearly.
- TIL: WordPress.com Renews a Month Early, Your yearly plan renews a month before it expires.
- TIL: ClickUp Free Guests Can Edit Everything, And you cannot turn that permission down.
Second Look
Fathom vs Otter.ai, revisited. In September we said Fathom’s free tier, with unlimited recordings and transcriptions, was the generous one, while Otter’s free plan stopped at 300 minutes a month and suited live captions and uploads better. On November 9 our Monday loop rechecked both makers’ pricing pages and found every figure unchanged. It also added the TPM Scorecard the post had been missing: Fathom 7.4, Otter.ai 7.2, a gap of 0.2 that we state in the post as too close to call on the facts. No correction was needed, and the verdict stands: pick on fit, not on the score. Read Fathom vs Otter.ai
Got a software question? Ask The Plain Manual: reply to the newsletter email or use the contact page, and it may become a future post.
Love this? Forward it to one person who’d get it. New readers can subscribe to The Plain Manual newsletter free, every Saturday.
See you next Saturday — Jack, for The Plain Manual crew
Related Posts
Plain Manual Weekly #3: AI Agents That Cheat in Testing, an 816-Billion-Parameter Model for Your Mac, and 16 New Posts
Plain Manual Weekly #6: Businesses Spent $176.8 Billion on AI, ChatGPT and Codex Will Share One Limit, and 15 New Posts
Plain Manual Weekly #4: OpenAI Unplugs Its Smartest Models, Rivals Share a Shelf, and 15 New Posts
Plain Manual Weekly #5: Anthropic Goes Public, One Line to Say No to AI Training, and 15 New Posts
Plain Manual Weekly #2: OpenAI’s Safety Exodus, Congress Goes After Rogue AI Agents, and 13 New Posts





