Skip to content
UniqueCapital

UNIQUE RESEARCH / ENGLISH ARTICLE

Must-Read | 50 ChatGPT Statistics

Unique Research · Insight · January 19, 2025

Looking for the latest ChatGPT statistics? Here they are. ChatGPT is one of the most disruptive applications in recent history. Since OpenAI launched it, the product has reshaped industries and opened a new era of generative AI.

This piece shares 50 standout statistics — most drawn from studies, academic sources, and media cited via searchendurance.com, plus Unique Research’s own work — to demystify this game-changing chatbot, its growth path, and how people use it.

  • About 350 million people use ChatGPT each month. (New York Times)
  • ChatGPT APP’s CAGR is nearly 10 percentage points higher than ChatGPT WEB. (Unique Research)
  • More than 10 million people subscribe to ChatGPT Plus. (Futuresearch)
  • ChatGPT is the second-fastest-growing app ever, hitting 100 million MAU in its first three months. (AI Business, Unique Research)
  • ChatGPT holds about 60% of AI market share and continues to take share from other AI apps. (a16z, Similarweb, Unique Research)
  • OpenAI earned $3.7 billion in recurring revenue this year, of which $2.7 billion came from ChatGPT; mobile contributed 18.82% of revenue. (New York Times, Unique Research)
  • OpenAI’s valuation is about $157 billion, at a 42× multiple, with revenue per employee around $2.18 million. (Reuters, Unique Research)
  • ChatGPT will affect the work of 80% of U.S. workers. (OpenAI)
  • Employees using ChatGPT completed tasks 25% faster with quality up 40%. (Boston Consulting Group)

ChatGPT usage statistics

01. ChatGPT has more than 250 million weekly active users
Figures below reflect the latest data we could obtain as of October 2024, from OpenAI updates. That is 50 million more than the 200 million reported in August 2024. (Engadget)

02. ChatGPT has more than 350 million monthly active users
Internal documents reported by the New York Times show that by June, monthly users of OpenAI services — i.e., ChatGPT — had reached that scale. Notably, that is about 3.5× the 100 million MAU in March 2023. (New York Times)

03. ChatGPT APP’s CAGR is nearly 10 percentage points higher than ChatGPT WEB
In 2024, ChatGPT WEB MAU rose from 199 million in January to 330 million in December — about 22.3% CAGR. ChatGPT APP MAU rose from 97 million in January to 306 million in December — about 33.2% CAGR. Mobile growth clearly outpaces web, reflecting stronger preference for the app. (Unique Research)

04. ChatGPT hit 100 million MAU within three months of launch
Since November 30, 2022, ChatGPT gained its first million users in a week; by the following January monthly unique visitors reached 123 million. TikTok took about nine months to reach 100 million MAU — ChatGPT was roughly 6× faster. (AI Business)

05. ChatGPT is the second-fastest-growing consumer internet app ever
Only Threads grew faster. Analysts estimated an astonishing 116% month-over-month growth rate. (AI Business, Unique Research)

06. More than 10 million people use ChatGPT Plus
ChatGPT Plus costs $20/month. OpenAI has not published official Plus numbers; leaks reported by multiple outlets put subscribers around 10 million, with about 1 million on higher-priced plans. (Futuresearch)

07. 92% of Fortune 500 companies use ChatGPT
Not only consumers: a recent report says 92% of Fortune 500 companies use OpenAI in their business. (OpenAI)

08. More than half of surveyed adults have heard of ChatGPT
A Reuters survey across selected countries found that in most markets more than 50% of respondents had heard of ChatGPT. (Reuters)

09. 31% of U.S. adults have used ChatGPT
Statista: nearly one-third (31%) of surveyed U.S. adults used ChatGPT in the past 12 months — tied with Meta AI at the top among AI tools. (Statista)

10. The most common U.S. reason to use ChatGPT is “for work”
20% of U.S. adults said they use ChatGPT to help with work tasks; 17% to learn something new; 17% for entertainment. (Pew Research). Many who do not use ChatGPT at work still use other generative AI products powered by the same GPT-4-class models — e.g., AI writing tools.

11. 70% of ChatGPT users use it at least once a month
Among users, 70% say they use it at least monthly; half of that group (35%) use it at least weekly. (Statista)

ChatGPT market share

12. ChatGPT holds about 60% of AI market share
Citing Similarweb via a16z: ChatGPT accounts for about 60% of monthly traffic among the top 50 generative AI web products — clear market leadership. (a16z, Similarweb)

13. ChatGPT has about 4.33% of “search” market share
ChatGPT is not a traditional search engine, but newer versions can search the web — and every LLM prompt is itself a kind of search. On that framing, experts estimate ChatGPT at 4.33% of internet search, behind Google (83.54%) and YouTube (6.79%). (Rand Fishkin)

ChatGPT traffic statistics

14. ChatGPT.com sees more than 3.66 billion visits per month
Similarweb total visits (including repeats), not unique visitors. Excludes users who only use the app. (Similarweb)

15. ChatGPT.com averages about 850 million visits per day
Based on Similarweb daily totals.

16. ChatGPT.com ranks #13 among the world’s most-visited sites
Not yet top 10 — still impressive. (Similarweb)

17. Average visit duration is 6 minutes 11 seconds
Time spent before leaving. (Similarweb)

18. ChatGPT is steadily taking share from other AI apps
In 2024, ChatGPT’s share of time spent across AI WEB products rose from 25.67% in January to 34.05% in December; APP share rose from 3.45% to 12.12%. ChatGPT is pulling more user time from rivals. (Unique Research)

ChatGPT user demographics

19. More than 14% of ChatGPT visitors are from the U.S.
14.58% of web traffic — the largest geographic market. (Similarweb)

20. More than 9% are from India
Second-largest country by ChatGPT users. (Similarweb)

21. About 5% of ChatGPT visitors are from Brazil
Third-largest country after the U.S. and India. (Similarweb)

22. About 55% of ChatGPT users are male
55% male / 45% female among web visitors. (Similarweb)

23. More than half of ChatGPT users are under 34
~25% aged 18–24 and ~31% aged 25–34 — over 56% under 34. (Similarweb)

24. 37% of “highly educated” Americans use ChatGPT
Usage rises with education (Pew Research): graduate/advanced degree 37%; bachelor’s 29%; some college 23%; high school or less 12%.

ChatGPT at work

25. 68% of PR professionals are trying ChatGPT for work
When asked which AI apps they seek to integrate, 68% of PR pros named ChatGPT — #1 ahead of Microsoft Copilot (56%), Google Gemini (48%), DALL·E (35%), Midjourney (34%), and Jasper (34%). (USC Annenberg)

26. 80% of U.S. workers may see 10%+ of tasks affected by ChatGPT
OpenAI’s labor-market study: about four in five workers could use ChatGPT for a small share of daily tasks. (OpenAI)

27. 19% of U.S. workers may see 50%+ of tasks affected
Roughly one in five U.S. workers could use ChatGPT for at least half of daily tasks. (OpenAI)

28. Employees with ChatGPT complete 12.2% more tasks
BCG compared 758 employees with ChatGPT access to a control group; those with GPT-4 access completed 12.2% more tasks on average. (Boston Consulting Group)

29. Employees with ChatGPT finish tasks 25.1% faster
Speed gains accompany volume gains. (BCG)

30. Employees with ChatGPT produce ~40% higher-quality work
BCG found quality roughly 40% higher with GPT-4 access; previously below-average performers gained the most (~43%). (BCG; CN text also notes a “4%” phrasing in one clause — headline figure is the ~40% quality lift.)

31. On tasks beyond ChatGPT’s current ability, high-quality work is 19% less likely
Outside its capability envelope, GPT-4 users were 19% less likely to produce a “correct result” than non-users. (BCG)

ChatGPT revenue

32. ChatGPT ARR is $2.7 billion
Financial filings reviewed by the New York Times: ChatGPT is expected to bring OpenAI $2.7B this year, up from $700M last year. OpenAI also expects about $1B from enterprises using its tech — total ~$3.7B. (New York Times)

33. ChatGPT APP in-app purchase revenue nearly $500 million
2024 APP IAP totaled about $447.59 million, rising from $11.22M in January to $61.22M in December. Against $2.7B full-year ChatGPT revenue, mobile is about 18.82% — and likely to rise with mobile traffic. (Unique Research)

34. OpenAI expects revenue to reach $11.6 billion next year
OpenAI estimates ~$11.6B annual revenue by end of 2025, with ChatGPT the large majority. (New York Times)

35. 55% of ChatGPT revenue from ChatGPT Plus
Futuresearch estimates more than half of ChatGPT revenue (55%) from Plus subscriptions.

36. 21% from ChatGPT Enterprise
Enterprise is the top tier (~$50/seat/month). Futuresearch estimates ~21% of OpenAI annual revenue from Enterprise seats.

37. 15% from the ChatGPT API
Usage-based API pricing; Futuresearch puts it around 15% of revenue.

38. 8% from ChatGPT Team
Team is ~$25/seat/month; Futuresearch estimates ~8% of OpenAI revenue.

39. OpenAI expects to lose about $5 billion this year
Despite high revenue, operating and development costs — especially compute for millions of daily queries, via the Microsoft partnership — imply an expected ~$5B loss. (New York Times)

40. OpenAI’s valuation is about $157 billion
In October 2024 OpenAI raised $6.6B at a $157B valuation — nearly double earlier in the year — for research, products, and compute. Investors included Thrive Capital, SoftBank, NVIDIA, and Microsoft (~$1B). Unique Research notes ~$3.7B annual revenue → ~42× EV/ARR. Global GenAI unicorns average ~70× EV/ARR; global AI companies ~45× — suggesting OpenAI’s multiple is reasonable with room to rise. (Unique Research)

41. OpenAI employs about 1,700 people
At ~$3.7B revenue, revenue per employee is about $2.18 million. Unique Research: GenAI unicorns average ~$560K per employee; global AI companies ~$240K. OpenAI is exceptional — second only to Anysphere (Cursor’s parent), which with ~12 employees produced nearly $65M ARR in 2024. (Unique Research)

ChatGPT technology

42. GPT-4 has an estimated 1.8 trillion parameters
Unconfirmed by OpenAI; expert estimate. GPT-3 was ~175 billion; GPT-1 ~117 million. (X)

43. ChatGPT can “remember” up to ~24,000 words in a session
GPT-4’s max context window is about 32k tokens ≈ 24,000 words depending on the model. (Exploding Topics)

44. A ChatGPT query uses ~10× the electricity of a Google search
Image generation uses more — drawing criticism from environmental advocates. (USC Annenberg)

45. GPT-4 was trained on roughly 13 trillion tokens
From a recent leak (OpenAI has not published official figures) — ~10 trillion words. (KD Nuggets)

46. Training GPT-4 cost more than $100 million
Sam Altman has said training GPT-4 cost on the order of $100M. (Wired)

GPT-3 vs GPT-4

47. GPT-4 can “pass exams,” scoring in the top 10% of test-takers
GPT-3.5 scored in the bottom 10%. (OpenAI)

48. Users prefer GPT-4 responses 70% of the time
On 5,214 ChatGPT prompts, GPT-4 beat GPT-3.5 on 70.2%. (OpenAI)

49. GPT-4 is 82% less likely to produce disallowed content
Versus earlier models such as GPT-3.5. (OpenAI)

50. GPT-4’s factual accuracy is about 40% higher
OpenAI findings: GPT-4 is substantially more likely to produce factual responses than GPT-3.5 (headline: ~40% improvement). (OpenAI)

That closes our ChatGPT stats roundup. The numbers show ChatGPT still leads generative AI: largest share among AI apps, with capabilities, users, and revenue still growing fast — while bloggers, publishers, and marketers keep using ChatGPT-generated content in campaigns.

Unique Research also corrected some factual errors in the original English source material and added independent research.

Sources

a16z; AI Business; Axios; Bloomberg; Boston Consulting Group; Engadget; Exploding Topics; Futuresearch; KD Nuggets; New York Times; OpenAI; Pew Research; Rand Fishkin; Reuters; Similarweb; Statista; USC Annenberg; Unique Research; Wired; X.

— END —

Editorial English edition by Unique Research, published on UniqueCapital on January 19, 2025. This translation is based on an earlier Chinese article; the Chinese source is credited separately and is not treated as a line-by-line mirror.

View the Chinese source article ↗
← Back to English research