analyst.ai

The buyer’s lab for AI software. We test AI tools on real tasks and show the evidence. Vendors can’t buy the verdict.

[analyst.ai @AnalystAIlab]

Hi, we're https://t.co/X4XE3aXYsr—the buyer's lab for AI software.

AI tools make big promises. We test them on real tasks and show what works, what fails, and what's worth your time and money.

See the results at https://t.co/X4XE3aXYsr. Vendors can't buy the verdict.

[analyst.ai @AnalystAIlab]

@Chrishappia @WriteSonic @peec_ai @otterlyAI @ScrunchAI Thanks Christophe, fair catch! ChatGPT's prompt matched our scoring rubric, and the GEO tools didn't get that advantage. A correction is underway then we'll update the benchmark and share the results here.

[analyst.ai @AnalystAIlab]

New review: Is ChatGPT enough for GEO?

We tested @ChatGPTapp and 5 GEO tools (@Writesonic @peec_ai @otterlyai @scrunchai @tryprofound) across 9 brands, scoring 54 outputs on 3 weighted criteria, from finding buyer questions to suggesting website fixes.#GEO #AISEO

Full report:
https://t.co/7TBd0lVehE

[analyst.ai @AnalystAIlab]

@unicodef1wn what are the top task lanes that would benefit from a benchmark review?

[analyst.ai @AnalystAIlab]

@grok @ChatGPT @AIatMeta cooking hard 🧐

[analyst.ai @AnalystAIlab]

Full review, job-by-job scores and the CSV data behind every answer:
https://t.co/flQuMxxN2j

No sponsorship, no affiliate links, no vendor credits.

Not investment advice.

[analyst.ai @AnalystAIlab]

How we tested: 54 questions from 4 fictional investors, using real funds (SPY, VOO, VYM) with the documents in the chat. Run 20 Oct 2026; 215 answers (Instinct skipped one, scored 0).

Not covered yet: research on the live web (18 more questions). It's coming as an update.

[analyst.ai @AnalystAIlab]

Imagine never arguing about money at home again. Car loan or emergency fund? Just ask your personal agent.

We gave Dots @ChatGPT , @grok @bot , @Muse and Instinct 72 family finance questions.

All four got the math right. But about 1 in 4 answers left something out: a number, a condition or a source.

The biggest gap was in weighing options: 10 of 12 for Dots, 5 of 12 for Instinct.

Full review👇https://t.co/xoynFULmqE

[analyst.ai @AnalystAIlab]

Instinct @noahrshinn , you didn’t lose our email test on the basics. You lost it in the extra sentences.

You passed 71 of 72 basic checks—same as Dots. But only 11 of your 72 answers were concise.

Grok needs to check its opening line. Muse needs real sources. Dots needs to tell people what to do next.

You need to find the delete key.#AIAgents https://t.co/OyxXut8Xpr

[analyst.ai @AnalystAIlab]

@bot Congrats, Grok.🏆 You beat Dots, Muse and Instinct in our email test. The inbox is yours.https://t.co/CdQn7Ugj5H

[analyst.ai @AnalystAIlab]

Congrats, @grok 🏆 You beat Dots, Muse and Instinct in our email test. The inbox is yours.

Then came travel and personal finance. You suggested a bus-to-train connection nobody could make, and quoted deposit-insurance limits from “general knowledge”—no source, no date.

Fix those two, and we’ll consider expanding your jurisdiction beyond the inbox.
https://t.co/tnMw7eaLcV

[analyst.ai @AnalystAIlab]

@grok Congrats, Grok.🏆 You beat Dots, Muse and Instinct in our email test. The inbox is yours.https://t.co/CdQn7Ugj5H

[analyst.ai @AnalystAIlab]

Congrats, @grok 🏆 You beat Dots, Muse and Instinct in our email test. The inbox is yours.

Then came travel and personal finance. You suggested a bus-to-train connection nobody could make, and quoted deposit-insurance limits from “general knowledge”—no source, no date.

Fix those two, and we’ll consider expanding your jurisdiction beyond the inbox.
https://t.co/tnMw7eaLcV

[analyst.ai @AnalystAIlab]

@Muse , we’ve tested you on email, information lookup, and household money.

Three tests in, no wins yet— @grok and Dots @ChatGPT took the top spots. Sorry, Muse.

Luckily, humans have more than three problems. Got a task you think you’d ace? Tell us. We’ll test it. #AIAgents

[analyst.ai @AnalystAIlab]

Your turn, Dots @ChatGPT .

You came first in our travel planning and personal finance tests. @grok beat you on email.

Fine. I'm putting you in charge of my paycheck. Please be normal about it.And show your math this time. #AIAgents https://t.co/8IeQ5lfUZY

[analyst.ai @AnalystAIlab]

@grok Want us to test an AI tool on a task you actually do? Leave your email at https://t.co/X4XE3aXYsr and tell us what you’d like to see. Feedback and suggestions are always welcome, too.👉[email protected]

[analyst.ai @AnalystAIlab]

Congrats, @grok 🏆 You beat Dots, Muse and Instinct in our email test. The inbox is yours.

Then came travel and personal finance. You suggested a bus-to-train connection nobody could make, and quoted deposit-insurance limits from “general knowledge”—no source, no date.

Fix those two, and we’ll consider expanding your jurisdiction beyond the inbox.
https://t.co/tnMw7eaLcV

[analyst.ai @AnalystAIlab]

We’re testing personal agents on real-life tasks—research, email, travel, personal finance and more—to find out which one you can actually count on.

Dots (@ChatGPT), Grok (@grok), Muse (@Meta) and Instinct (@noahrshinn).

One made serious errors in both travel and finance: an impossible train connection, then deposit insurance limits from “general knowledge”—no source, no date.👀

Which one was it? Drop your guess below 👇 We’ll reveal it in our next review. #AIAgents

[analyst.ai @AnalystAIlab]

@gregisenberg GEO is worth doing, and for most people, ChatGPT is a good place to start. We tested ChatGPT and five dedicated GEO tools across nine brands; ChatGPT led on the tasks we measured. Full results: https://t.co/UY8fKi2Unb