The applied-AI desk for people who run things.
Two tools that get shortlisted against each other, the prices we have recorded for both, and a verdict on which to pick and when. We write one of these only when the decision is real.
Which assistant should a small team pay for?
Both have a free tier good enough to trial for a fortnight. Pay for the one where your second-most-common task also works, because you will not buy two.
ChatGPT $20/mo / seat, per seat · Claude $17/mo
Long-document work: Claude or Gemini?
Judge them on your own worst document, not on a demo. Longer context only helps if the model does not quietly lose the middle of it.
Claude $17/mo · Gemini $19.99/mo / seat, per seat
Where should a developer's AI budget go?
Copilot is a better colleague; Cursor is a better pair. Pick on whether your bottleneck is writing lines or understanding how the lines connect.
GitHub Copilot $10/mo / seat, per seat · Cursor $20/mo / seat, per seat
Is paying for AI inside your notes worth it?
Notion AI is a convenience purchase. Buy it when context-switching costs you more than the subscription does.
Notion AI $10/mo / seat, per seat · ChatGPT $20/mo / seat, per seat
Making images for a publication: which one?
These are not rivals so much as different jobs. Buying both is defensible; buying neither and hoping is not.
Midjourney $10/mo / seat, per seat · Canva $15/mo / seat, per seat
Research with sources, or one assistant for everything?
If your work is ever questioned, pay for the one that shows its working.
Perplexity $20/mo / seat, per seat · ChatGPT $20/mo / seat, per seat
Automating a small business: which platform?
Zapier's catalogue is the moat. Make's price is the argument. Count your integrations before you count your budget.
Zapier $19.99/mo / seat, per seat · Make $9/mo / seat, per seat
Copy editing prose you have already written
Grammarly fixes sentences. A chat model replaces them. Which you need depends on whether the thinking is already done.
Grammarly $12/mo · ChatGPT $20/mo / seat, per seat
We could generate a page for every pair of tools we track, and each one would be a price table with a headline on top. Those pages help nobody — not the reader deciding between two products, and not the archive, which would fill with near-identical pages. A comparison appears here when the two names genuinely compete for the same budget, and it carries a written reason rather than an automated score.
The day's consequential AI developments with the operational consequence stated, plus every price change we detect. Free, one send a day, one click to leave.
No third parties, no sponsored placements inside the brief, no list rental.