aiintelligence

Tools / chat

Claude

There is a free versionconfirmed on Claude’s own pricing pageclaude.com/pricing8 Aug 2026
App Store rating
4.72★
Apple App Store (iTunes Lookup API)8 Aug 2026
From
$20/mo
claude.com/pricing8 Aug 2026
2 plans verified · all tiers ↗
Battle record
1 tested
1 too close to call
What people say
Mostly criticised
4 of 29 posts were positive (14%) · 95% range 5–31%
Partial coverage this run: reddit partial, youtube timeout. The sample is what those sources returned, not everything that exists.
“YSK: A Whole Bunch of People’s Claude Chats Are Publicly Accessible Online, and There’s Some Wildly Private Stuff in There”
on Claude · num_comments 598, score 14960
73 public posts sampled over 90 days; 29 carried a clear view, 44 were announcements or neutral and are excluded from the score. Sources: github (ok), hackernews (ok), reddit (partial), youtube (timeout). Classified by google/gemini-3.1-pro-preview under method audience-2026-08-c. Public posts skew negative — people write when something breaks — so this compares like with like rather than rating quality in the absolute. How this is measured
Visit Claude ↗

Is it safe?

Four things worth knowing before you type anything into Claude. Where Claude says it itself, we quote them and link to the page we read it on.

1

What happens to what you type?

They use your conversations to train their AI models, but you can turn this off in your account settings.

“We may use your Inputs and Outputs to train and improve Anthropic AI models, unless you opt out through your account settings.”

their own words — anthropic.com ↗ · read 7 Aug 2026

They keep your data for as long as they decide it is necessary for their purposes.

“Anthropic retains your personal data for as long as reasonably necessary for the purposes and criteria outlined in this Privacy Policy”

their own words — anthropic.com ↗ · read 7 Aug 2026
2

Can it be wrong?

The company warns that its AI can make mistakes and tells users to independently check the accuracy of the results.

“You should not rely on any Outputs or Actions without independently confirming their accuracy.”

their own words — anthropic.com ↗ · read 7 Aug 2026
3

What should you never type into it?

Until you turn training off in the settings, treat anything you type as something a stranger could one day read. Either way: never passwords, bank details or someone else's private information.

This one is our advice, not theirs — it follows from what their policy says above.
4

Do you have to give them a card?

No — there is a free version you can use without paying. The paid plans are extra, and you choose to take one.

confirmed on Claude’s own pricing page — claude.com/pricing ↗ · read 8 Aug 2026

Every quoted line was found word-for-word in the document linked beside it. Nothing appears here from memory: if we could not find the sentence, the answer says so instead. How we check this

Battles fought
Claude vs ChatGPT · Writing5–4
Every claim we hold (8)
The 29 posts that carried an opinion (29)

We found 73 public posts about Claude. 44 of them carried no opinion at all — release notes, tutorials, announcements, and posts that only mention it in passing — so they are excluded rather than counted as half a vote. These are the 29 that took a side, with the classifier’s reason beside each. Open any of them and disagree with us.