Revunda · Free mini check · Sample reports · Rating badge · All app reports · App comparisons
DeepSeek app reviews: what users complain about
We ran Revunda on 300 of the most recent public App Store reviews of DeepSeek - AI Assistant (AI assistant, by 杭州深度求索人工智能基础技术研究有限公司) from 3 storefronts, dated 2025-05-13 to 2026-10-22. Average rating in this sample: 3.01★. Its all-time US App Store rating is 3.9★ from 12K ratings. 118 reviews contained a complaint; the biggest theme was bugs (78%).
Free App Store check Get the full report for your app – €19 20 ready replies to your real bad reviews – €9 See a sample report
Competing with DeepSeek? Compare your app with DeepSeek and one more rival – Competitor Pack €29 · see a real replies sample
Download the full DeepSeek report (PDF, free)
Complaint categories
| Category | Reviews | Share of complaints | Avg severity (0–3) |
|---|---|---|---|
| Bugs | 92 | 78% | 2.2 |
| Feature requests | 51 | 43% | 2.0 |
| UX / design | 35 | 30% | 2.0 |
| Performance | 25 | 21% | 2.3 |
| Crashes / freezes | 10 | 8% | 2.5 |
| Pricing / subscription | 4 | 3% | 2.6 |
A review can count in several categories, so shares add up to more than 100%. Severity is the AI's estimate of how bad the problem is for the user.
Top issues, ranked by reviews × severity
- Improve content quality: 64 complaining reviews, avg severity 2.3/3. What to do: Review the content areas users name; add a "report a problem" button on content.
“Gives wrong answers very often!” (1★, v1.3.3, CA) - Soften the redesign backlash: 65 complaining reviews, avg severity 2.2/3. What to do: Explain the change in-app, restore the most-missed option, or offer a setting for the old behaviour.
“You need to be fired and never find work for the rest of your worthless life” (1★, v1.8.2, CA) - Ship the most requested feature: 41 complaining reviews, avg severity 2.0/3. What to do: Read the quotes below for the exact request; consider a public roadmap.
“After the latest update, the app has become almost unusable for roleplay. Characters are emotionless, forget the context, show no initiative, and the responses have turned into short, lifeless, generic text. Please bring back the previous quality and modes — t…” (1★, v2.5.2, CA) - Rethink usage limits: 22 complaining reviews, avg severity 2.2/3. What to do: Users read hard limits as a cash grab; consider softer limits or earning more through use.
“The limit on editing has become very stale, so it has also become very unintelligent. You've ruined a good app, there's nothing more to say. It will be the same as with the charterAi.” (1★, v2.6.2, US) - Fix login / account access: 10 complaining reviews, avg severity 2.4/3. What to do: Audit the sign-in flow (SSO, password reset, verification emails) on the latest version; add clear error messages.
“The recent update completely lobotomized this AI: - It has trouble with memories even in short chat logs - It hallucinates, making up false information - Doesn’t follow half of the prompt - Strict regeneration limit/timeout which is especially bad because it…” (1★, v2.6.1, US) - Fix features broken by the latest update: 10 complaining reviews, avg severity 2.2/3. What to do: Add regression tests for the features users name in these reviews before the next release.
“Everything used to work perfectly: when sending a photo through the "plus" button, it was attached to the current dialog. But today, when trying to send a photo, a message appears: "This message will be sent to the new chat." After that, a new dialogue is cre…” (5★, v1.8.1, CA)
Rating by app version (newest first)
| Version | Reviews | Avg stars | Share with a complaint |
|---|---|---|---|
| 2.7.0 | 4 | 1.0 | 75% |
| 2.6.3 | 21 | 3.7 | 29% |
| 2.6.2 | 57 | 2.6 | 47% |
| 2.6.1 | 30 | 2.3 | 57% |
| 2.5.2 | 12 | 1.8 | 50% |
Versions with fewer than 3 reviews in the sample are left out.
FAQ
What do users complain about most in DeepSeek?
In 300 recent App Store reviews (2025-05-13 to 2026-10-22), the most common complaint theme was bugs (78% of complaining reviews), and the top specific issue was “improve content quality” (64 reviews).
How many DeepSeek reviews are negative?
118 of 300 recent reviews (39%) contained a complaint. Average rating in this sample: 3.01★. Its all-time US App Store rating is 3.9★ from 12K ratings.
Can I get the same report for my own app?
Yes. Paste your App Store or Google Play link at revunda.xyz/apps/ and you get the full PDF (up to 300 recent reviews, 7 categories, 18 specific issues, version trend, 10 ranked fixes) within 24–48 hours. A free mini check shows the top 3 themes in seconds.
How to read this
Apple's public feed only shows recent reviews, so this is a snapshot of the last weeks or months, not the app's full history. Popular apps attract complaints in proportion to their size; the useful part is which issues repeat. Generated 22 October 2026. We have no relationship with 杭州深度求索人工智能基础技术研究有限公司.
Compare DeepSeek with rivals
- ChatGPT vs DeepSeek: which has fewer problems?
- DeepSeek vs ChatGPT vs Claude: which has fewer problems?
More app complaint reports
- ChatGPT reviews: top complaints (AI assistant)
- Character AI reviews: top complaints (AI assistant)
- Google Gemini reviews: top complaints (AI assistant)
- Claude reviews: top complaints (AI assistant)
- Perplexity reviews: top complaints (AI assistant)
- Microsoft Copilot reviews: top complaints (AI assistant)
- OkCupid reviews: top complaints (Dating)
- YouTube Music reviews: top complaints (Music)
All 216 app reports · Free App Store rating badge for your site
Free App Store check Get my report – €19 20 ready replies to your real bad reviews – €9 See a sample report
Competing with an app like this one? The Competitor Pack (€29) compares your app with up to 2 competitors side by side.