Investigate accuracy firstextreme
6.3% here · 3.6% in its market · 1.5% store-wide.
Claude by Anthropic
Anthropic PBC
4.72star252,861 ratings
5 decisions.
Shares of this app’s 686 recent negative reviews.
Billing is at typical. Skip it
24.9% here · 44.5% in its market · 30.8% store-wide. Normal for the store.
Watch limits & capsextreme
7.7% here · 16% in its market · 1.9% store-wide.
Skip performance, login & auth and sync & data work
Performance, Login & auth, Sync & data all run at or below the store’s typical rate.
Protect ads, UX and support
Ads, UX, Support all run well below the store’s typical rate.
n = 60. Too thin to act on. We recount weekly.
Read those two together. Ratings are cumulative, reviews are recent. A 4.72star average beside 61% negative is the finding, not an error.
Ratings.
254,232 ratings · Apple’s own count, fetched daily
5star225,230
4star11,124
3star4,291
2star2,512
1star11,075
The market in numbers
All 8 panel members · 2026-09-02 snapshot.
| App | Ratings | Accuracy |
|---|---|---|
 Perplexity - AI Search & Chat | 502,698 | 7.9% |
 Claude by Anthropic | 252,861 | 6.3% |
 DeepSeek - AI Assistant | 11,342 | 5.6% |
 AI Chatbot - Nova | 125,495 | 4.6% |
 Grok AI | 1,387,721 | 3% |
 RoboAI - Chat & Ask AI Chatbot | 16,754 | 2.2% |
 Chatbot AI Assistant - Genie | 200,172 | 1.3% |
 Lloyd - Your AI Superpower | 23,588 | 0.6% |
Counted from real App Store reviews · 2026-09-02
How this is counted
Every number is arithmetic over the reviews we hold, re-counted weekly: nothing modelled, nothing scored. Themes come from a frozen pattern classifier rather than a model reading reviews, so the same review always lands in the same bucket. Shares are of this app’s own recent negative reviews; the market figure is its rivals with this app excluded. The graded words have fixed cutoffs against the store’s typical rate: elevated from 1.15, high from 2, extreme from 3.5 times it. The page says what is broken and how it compares, never how to fix it, and never whether this app can win.