ChatGPT Hits 400 Million Weekly Users, OpenAI Reveals

ChatGPT statistics for 2026 reveal slowing user growth despite hitting 400 million weekly active users. Revenue, enterprise adoption, and what's next for OpenAI.

ChatGPT Statistics 2026: 400 Million Weekly Users and the Slowdown No One Saw Coming

Key Takeaways

- ChatGPT reached 400 million weekly active users in February 2026, up from 300 million in December 2025 — a 33% increase over 10 weeks, according to OpenAI CEO Sam Altman - Daily active users hit 100 million for the first time in January 2026, per internal company figures shared with The Information - Growth rate is decelerating sharply: Monthly user growth dropped from 15-20% in early 2025 to roughly 5-7% by late 2025, based on Sensor Tower app download data - Google Gemini now claims 350 million monthly users, up from 100 million in May 2024, representing faster relative growth from a smaller base, according to Google CEO Sundar Pichai's February 2026 earnings call - Anthropic's Claude has 15 million monthly users as of January 2026, per The Information — small but growing 40% quarter-over-quarter - Enterprise AI spending will reach $151 billion in 2026, up from $67 billion in 2024, according to Gartner's January forecast - OpenAI's annualized revenue hit $5 billion by late 2025, but losses exceeded $5 billion for the year, per The Information's financial reporting

---

User Growth: The Numbers Behind the Milestone

OpenAI's announcement of 400 million weekly active users marks a watershed moment for consumer AI adoption. But the trajectory tells a more complicated story than the headline suggests.

The Growth Curve Is Flattening

ChatGPT added its first 100 million users in roughly 2 months after its November 2022 launch — the fastest-growing consumer application in history at that time, according to UBS analysis. The next 100 million took approximately 9 months. The 300 million to 400 million jump required 10 weeks — faster in absolute terms, but representing a much slower growth rate relative to the installed base.

Sensor Tower data shows ChatGPT mobile app downloads peaked in May 2023 with approximately 58 million monthly downloads globally. By December 2025, that figure had fallen to roughly 23 million monthly downloads — still substantial, but down 60% from the peak. Web traffic tells a similar story: SimilarWeb reports ChatGPT.com visits grew just 8% month-over-month in January 2026, compared to 25-35% monthly growth throughout 2023.

Key monthly active user milestones: - 100 million: January 2023 (2 months post-launch) - 200 million: August 2023 (9 months) - 300 million: December 2025 (26 months) - 400 million: February 2026 (28 months)

The acceleration from 300 to 400 million reflects OpenAI's aggressive holiday 2025 marketing push and the 12 Days of OpenAI event series, which drove 50 million new signups in December alone, according to company statements. Without that artificial spike, the underlying growth rate would likely have appeared even more modest.

Geographic Distribution: Where the Users Are

OpenAI has remained selective about geographic breakdowns, but third-party data fills some gaps. App Annie (data.ai) estimates the United States represents 18% of ChatGPT's mobile user base, followed by India at 12%, Brazil at 7%, and the UK at 5%. The remaining 58% is distributed across 150+ countries.

Notably, China is effectively absent from these figures due to access restrictions, though VPN usage likely contributes some unmeasured volume. Russia, Iran, and North Korea are blocked entirely per OpenAI's terms of service.

RegionShare of UsersEstimated MAU North America22%88 million Europe19%76 million Asia-Pacific (ex-China)31%124 million Latin America15%60 million Middle East & Africa9%36 million Other/Unknown4%16 million

Source: data.ai estimates, December 2025; OpenAI MAU figures

---

Competitive Landscape: Google and Anthropic Gain Ground

The 400 million milestone looks different when set against accelerating competition. OpenAI's first-mover advantage is eroding — not collapsing, but visibly thinning at the edges.

Google Gemini's Surge

Google's Gemini has emerged as the most credible threat to ChatGPT's dominance. 350 million monthly active users as of February 2026 represents 250% growth from the 100 million reported in May 2024, according to Pichai's statements. The integration strategy is working: Gemini is now the default assistant on 2.5 billion Android devices*, and Google reports "hundreds of millions" of users* have interacted with Gemini through Search, Docs, and Gmail without necessarily opening the standalone app.

Gemini's distribution advantages: - Default on Android: 2.5 billion devices globally - Google One AI Premium: 100 million subscribers as of October 2025, per Alphabet earnings - Workspace integration: 3 billion Google Workspace users have Gemini access through enterprise licenses

Anthropic's Niche Expansion

Anthropic's Claude remains a distant third in raw user numbers — 15 million monthly active users according to The Information's January 2026 reporting — but exhibits the highest growth rate among major competitors. The 40% quarter-over-quarter growth figure outpaces both OpenAI and Google on a percentage basis, albeit from a much smaller foundation.

Claude's user base skews heavily toward technical professionals and enterprises. Anthropic reports "tens of thousands" of enterprise customers, with particular strength in financial services, legal, and healthcare — sectors where Claude's emphasis on safety documentation and reduced hallucination rates resonates, per company marketing materials.

Competitive positioning by user base: PlatformMonthly UsersWeekly UsersGrowth Rate (QoQ)Primary Use Case ChatGPT500 million (est.)400 million8%General consumer Gemini350 millionNot disclosed15%Integrated productivity Claude15 millionNot disclosed40%Enterprise/technical Perplexity20 millionNot disclosed25%Search/research Copilot210 million (licensed)Not disclosed12%Microsoft ecosystem

Sources: Company disclosures, The Information, Sensor Tower estimates; figures as of February 2026

---

Enterprise Adoption: Where the Money Lives

Consumer user numbers generate headlines, but enterprise revenue determines viability. OpenAI's financial trajectory depends disproportionately on converting free users to paid tiers and expanding business contracts.

Revenue Breakdown: Consumer vs. Enterprise

OpenAI's $5 billion annualized revenue run rate (as of late 2025) splits roughly 60% consumer, 40% enterprise, according to The Information's financial reporting. The consumer segment is dominated by ChatGPT Plus subscriptions at $20/month — estimated at 15-20 million subscribers based on revenue figures and pricing. ChatGPT Pro at $200/month contributes a smaller but growing share, particularly among power users and developers.

The enterprise segment includes: - ChatGPT Enterprise: "Hundreds of thousands" of seats deployed, per OpenAI's November 2025 announcement - ChatGPT Team: $25-30 per user/month, targeting smaller organizations - API usage: Estimated $1.5-2 billion annual run rate, serving developers and third-party applications

OpenAI revenue composition (estimated, annualized): SegmentEstimated RevenueShare ChatGPT Plus$3.6-4.8 billion72-96% of consumer ChatGPT Pro$400-600 million8-12% of consumer ChatGPT Enterprise/Team$1.2-1.5 billion60-75% of enterprise API$1.5-2.0 billion25-40% of enterprise Total$5.0 billion100%

Sources: The Information, OpenAI statements, analyst estimates

Enterprise Spending Patterns

Gartner's January 2026 forecast projects $151 billion in global enterprise AI spending for 2026, representing 125% growth from $67 billion in 2024. The research firm notes that 67% of enterprises now have "active AI pilots or deployments," up from 23% in 2023.

But deployment doesn't equal satisfaction. McKinsey's December 2025 survey of 1,500 enterprise technology leaders found that only 31% of AI projects have moved beyond pilot to production at scale. The most common barriers: data quality issues (cited by 54%), integration complexity (47%), and cost unpredictability (41%).

---

Financial Performance: The Profitability Gap

OpenAI's user growth has not translated to financial sustainability. The company's $5 billion in 2025 losses — on $5 billion in revenue — reveals the structural economics of frontier AI development.

The Cost of Scale

Training and inference costs dominate OpenAI's expense structure. SemiAnalysis estimates GPT-4 class models cost $100-200 million to train, while next-generation models may exceed $1 billion per training run. Inference costs — the compute required to serve 400 million weekly users — scale linearly with usage.

Estimated cost structure (2025, annualized): - Compute/training: $3.0-3.5 billion (60-70% of costs) - Personnel: $800 million-1.0 billion (16-20%) - Other operating expenses: $700 million-1.2 billion (14-24%)

Microsoft's cloud infrastructure partnership provides some cost relief — OpenAI receives preferential pricing on Azure compute reportedly at 20-30% below standard rates, according to The Information. But the sheer volume of inference demand outstrips these discounts.

The Funding Imperative

OpenAI's January 2026 $40 billion funding round — the largest private tech financing in history — values the company at $300 billion, per Bloomberg reporting. The round was led by SoftBank's $15-25 billion commitment, with Microsoft, NVIDIA, and Abu Dhabi-based MGX participating.

The funding is structured in tranches contingent on OpenAI's conversion to a for-profit entity, a legally complex transition that has drawn regulatory scrutiny and lawsuits from co-founder Elon Musk. The company has until 2027 to complete the restructuring or face potential clawback provisions, according to term sheets reviewed by The Financial Times.

DateRoundAmountValuationLead Investor Jul 2019Series A$1 billionN/AMicrosoft Jan 2023Extension$10 billion$29 billionMicrosoft Feb 2024Tender$300 million$80 billionThrive Capital Oct 2024Series C$6.6 billion$157 billionThrive Capital Jan 2026Series D$40 billion$300 billionSoftBank

Sources: Company announcements, press reports, SEC filings

---

Usage Patterns: What People Actually Do With ChatGPT

User counts tell us reach; usage patterns reveal value. OpenAI has shared limited data on how the 400 million weekly users actually engage with the product.

Feature Adoption

OpenAI's December 2025 product update offered rare specifics: Canvas — the collaborative document editing interface — had been used by "tens of millions" of users since its October launch. o1 reasoning usage was described as "growing rapidly" without hard numbers. Custom GPTs in the GPT Store had generated "hundreds of millions" of interactions, with "thousands" of creators earning revenue through the revenue-sharing program.

Third-party research provides additional texture. A Pew Research Center survey of 5,000 U.S. adults in November 2025 found: - 72% of ChatGPT users primarily use it for "information lookup" (replacing search) - 45% use it for "creative writing or content generation" - 38% for "coding or technical assistance" - 23% for "learning or education" - 12% for "professional work tasks"

The professional work figure is notably low given enterprise marketing emphasis, suggesting either underreporting (users may not classify work tasks as such) or genuine limitations in workplace penetration — a gap that educators have also noted as they adapt to AI-generated content in academic settings.

Session Frequency and Depth

SimilarWeb data indicates average session duration of 8.2 minutes on ChatGPT.com, down from 10.5 minutes in early 2024. Sessions per weekly active user average 4.3, suggesting most users engage sporadically rather than daily — consistent with OpenAI's disclosure of 100 million daily users against 400 million weekly.

---

What Does This Mean for AI's Future?

The 400 million user milestone arrives at an inflection point. OpenAI's growth is maturing, not stalling — but the competitive dynamics that defined 2023-2024 are shifting irreversibly.

The Commoditization Pressure

Model capabilities are converging. Benchmark scores from LMSYS Chatbot Arena show Gemini 2.0 Flash, Claude 3.5 Sonnet, and GPT-4o within 2-3 percentage points on most leaderboards — differences imperceptible to typical users. As models become interchangeable, distribution and integration become decisive advantages. Google's Android integration and Microsoft's Office embedding threaten to disintermediate the standalone chat interface that built OpenAI's brand.

This convergence has made direct comparisons more relevant than ever, as explored in our head-to-head testing of Claude and ChatGPT, which found performance gaps narrowing across most real-world tasks.

The Regulatory Variable

The European Union's AI Act enforcement began in February 2026, with prohibited practices provisions now active. OpenAI has not disclosed how many users or features are affected by compliance modifications, but geofencing of certain capabilities in regulated jurisdictions is ongoing. The Trump administration's January 2025 executive order on AI infrastructure has accelerated data center permitting but introduced uncertainty around export controls on model weights — a tension that could bifurcate global AI markets.

---

Methodology and Sources

This analysis aggregates data from the following primary sources:

- OpenAI corporate disclosures: User figures from CEO statements, product announcements, and blog posts (February 2026) - Alphabet/Google earnings calls: Gemini user figures from Q4 2025 and Q1 2026 transcripts - The Information: Financial reporting on OpenAI revenue, losses, and funding terms (2024-2026) - Sensor Tower / data.ai: Mobile app download and usage estimates - SimilarWeb: Web traffic analytics for ChatGPT.com and competitors - Gartner: "Forecast: Enterprise AI Software, Worldwide, 2022-2028" (January 2026) - McKinsey & Company: "The State of AI in 2025: Enterprise Adoption" (December 2025) - Deloitte: "2025 State of AI in the Enterprise" report - Pew Research Center: "Americans' Use of ChatGPT and AI Tools" survey (November 2025) - MIT Sloan Management Review: "AI ROI: The Reality Gap" (October 2025) - SemiAnalysis: AI infrastructure and training cost estimates (2024-2026) - LMSYS Chatbot Arena: Crowdsourced model evaluation leaderboard

Secondary sources include SEC filings, press releases, and verified social media statements from named executives. All figures are attributed to their original sources; estimates are clearly marked as such.

---

What to Watch

The next 12 months will test whether scale or specialization wins in consumer AI. Three signals merit particular attention:

First, OpenAI's Stargate infrastructure project — the $500 billion U.S. data center buildout announced with the Trump administration — will either secure decisive cost advantages or become a stranded asset if model efficiency improves faster than projected. The first $100 billion phase breaks ground in Texas in March 2026. Second, Apple's AI strategy remains the largest uncommitted variable. Siri's delayed "LLM-powered" relaunch — now expected late 2026 — could redirect 1.5 billion iPhone users toward or away from third-party AI assistants. Apple's partnership negotiations with OpenAI, Google, and Anthropic are reportedly ongoing and contentious, per Bloomberg. Third, the enterprise API market is fragmenting. Amazon's Nova models, Meta's Llama enterprise offerings, and open-source alternatives are pressuring pricing. Watch for OpenAI's API revenue growth rate — if it slows below 20% quarter-over-quarter, the core business model faces stress regardless of consumer user totals.

The 400 million milestone is a checkpoint, not a destination. The race now is for sustainable economics — and that finish line remains distant for everyone.

---