Key takeaways
Anthropic's third-quarter 2026 revenue was $17.0 billion, per its S-1 amendment filed November 2, up from $11.6 billion in Q2 and $4.7 billion in Q1. Nine-month revenue was $33.2 billion, against $4.6 billion for all of 2025.
Anthropic is not yet consistently profitable. It reported operating income of $974 million in Q2 and an operating loss of $567 million in Q3. The nine-month net loss was $104.0 billion, of which $94.3 billion was a non-cash fair-value change.
The S-1 shows an enormous compute bill. Per 9fin's analysis, Anthropic has committed to spend as much as $807 billion on chips, data centres and cloud capacity, at least $430 billion of it non-cancellable.
Anthropic filed publicly for an IPO on October 26. It plans to list on the Nasdaq Global Select Market under the symbol ANT. Its last private round, in May 2026, valued it at $965 billion.
Claude leads OpenAI in US business adoption. Ramp's September 2026 data shows 43.8% of US businesses paying for Anthropic against 40.6% for OpenAI. More than 1,900 customers spent $1 million or more over the trailing twelve months to September 30.
Anthropic still publishes no user count. Its S-1 reports customers and dollars, not people. First Page Sage estimates 312.4 million monthly active users in October 2026, and Similarweb has claude.ai at roughly 950 million visits a month.
Claude's share of AI chatbot web traffic is 9.6%, up from 2.3% a year earlier on Similarweb's measure, and its US usage share is 11.2% on First Page Sage's. On referral traffic it is still tiny, at 0.8% of AI referrals in August (BrightEdge).
Revenue is mostly usage-based. Consumption-based model usage was $13.1 billion, or 80% of first-half 2026 revenue, and subscriptions were $3.2 billion, or 20%.
Two years ago, Claude was the AI assistant people compared unfavourably to ChatGPT on consumer benchmarks. In 2026, Anthropic's revenue passed $17 billion in a single quarter while its consumer footprint stayed a fraction of ChatGPT's and Gemini's. Understanding that split, with enterprise and developer revenue running far ahead of consumer reach, is the key to reading every other number in this post. For brands, it also means ranking in ChatGPT and Google still matter more for direct consumer discovery, even as Claude becomes a larger part of how professionals research.
Here is a full breakdown of the numbers as of October 2026.
How many people use Claude?
Anthropic has never published a monthly active user count, and its S-1 amendment reports none. The filing gives customers and dollars instead. The figures in circulation are third-party estimates:
Monthly active users: First Page Sage estimates 312.4 million across Claude's web, desktop, mobile and Claude Code surfaces in October 2026. That is one vendor's model, not an Anthropic figure.
Website traffic: claude.ai draws roughly 950 to 969 million visits a month through August 2026, per Similarweb.
Share of AI web traffic: Claude held 9.6% in September 2026, up from 2.3% in October 2025, per Similarweb. Only ChatGPT (56.4%) and Gemini (25.0%) are larger.
Share of US chatbot usage: First Page Sage puts Claude at 11.2% in October 2026, up from 8.5% in January.
Claude's share over time
Month (2026) | Claude share of US chatbot usage (First Page Sage) |
|---|---|
January | 8.5% |
April | 8.2% |
June | 11.5% |
August | 10.2% |
September | 10.6% |
October | 11.2% |
Source: First Page Sage, Top Generative AI Chatbots by Market Share, November 2026. Claude's share rose in June and has held between about 10% and 11.5% since.
Revenue: Q3 2026 reached $17 billion
Anthropic's S-1 and its November 2 amendment give the first official view of its revenue by quarter. The 2026 interim figures are unaudited.
Period | Revenue |
|---|---|
2023 | $41 million |
2024 | $386 million |
2025 | $4.6 billion |
Q1 2026 | $4.7 billion |
Q2 2026 | $11.6 billion |
Q3 2026 | $17.0 billion |
Nine months of 2026 | $33.2 billion |
Source: Anthropic S-1 (October 26, 2026) and S-1/A (November 2, 2026), via 9fin and Mission Growth. In the first half of 2026, consumption-based model usage contributed $13.1 billion (80%) and subscriptions $3.2 billion (20%), per CNBC's reading of the S-1 via Rhino Rank. No single customer accounted for more than 10% of revenue, and the top five customers made up about 25%, per 9fin.
Anthropic also reports run-rate revenue, which the S-1 defines as the trailing 28 days of consumption fees multiplied by 13, plus monthly subscription fees multiplied by 12. It is not a substitute for reported revenue:
Date | Annualised run-rate revenue |
|---|---|
December 31, 2025 | About $9 billion |
February 12, 2026 | $14 billion |
April 6, 2026 | Over $30 billion |
May 2026 | Over $47 billion |
End of July 2026 | $65 billion (reported by CNBC, not published by Anthropic) |
Source: Mission Growth's compilation of Anthropic announcements, November 2026; CNBC and Reuters, August 17, 2026. The S-1/A leaves its October 31 run-rate blank. Q3 revenue multiplied by four is $68 billion, 5% above the July run rate, which suggests the run-rate figures are consistent with what Anthropic reported.
Anthropic also books a share of revenue gross when sales flow through cloud partners, which inflates its top line next to a company reporting net. OpenAI, by contrast, was at about $50 billion annualised at the end of September on its own accounting, per our ChatGPT statistics, so the two are not directly comparable.
Profitability and losses
Revenue growth has not produced steady profit yet. Per the S-1/A, Anthropic had operating income of $974 million in Q2 and an operating loss of $567 million in Q3, giving a nine-month operating loss of $1.3 billion. The first-half operating loss narrowed to $693 million from $2.1 billion a year earlier.
The headline net loss is far larger and mostly accounting. The nine-month net loss was $104.0 billion, but $94.3 billion of that was a non-cash fair-value change, and the first-half loss of $80.7 billion included a $72.4 billion non-cash charge from revaluing convertible notes as Anthropic's value rose.
What Anthropic has committed to spend on compute
The S-1 shows how much capacity Anthropic has contracted to serve demand. Per 9fin's analysis, it has committed to spend as much as $807 billion on chips, data centres and cloud computing in the coming years, at least $430 billion of which cannot be exited.
Commitment | Size and term |
|---|---|
$114.5 billion through July 2033 | |
Amazon | $110 billion through April 2036 |
Microsoft | $44.9 billion through May 2033 |
Broadcom (five-year equipment leases) | $161.2 billion, excluding interest |
SpaceX's AI subsidiary | Up to $120.5 billion |
Additional TPU capacity, 2028 to 2031 | $125 billion to $150 billion estimated |
AMD hardware (July agreement) | More than $20 billion |
Compute leases not yet started (end of June) | $321.4 billion |
Source: 9fin, Anthropic S-1 analysis, October 27, 2026. Compute and infrastructure was Anthropic's largest expense, at about $8.9 billion in the first half of 2026 against $2.3 billion a year earlier, per Reuters. The commitments dwarf current spending, which is why the S-1 is read as a bet that revenue keeps compounding.
Claude Code and spending per user
The S-1/A reports no Claude Code revenue or user total, but it shows how much the tool lifts spending. In Q3, Claude Enterprise users who did not use Claude Code spent less than $75 a month on average, while the top 5% who did spent about 20 times that amount.
How many businesses use Claude?
By multiple enterprise-spend measures, Anthropic is now ahead of OpenAI:
Share of US businesses paying (Ramp) | Anthropic | OpenAI |
|---|---|---|
August 2026 | 43.5% | 39.5% |
September 2026 | 43.8% | 40.6% |
Source: Ramp AI Index, September 2026 data. Ramp's index covers card spend across 85,000+ firms, and 55.9% of US businesses on Ramp paid for AI in September. OpenAI gained 1.1 points in September against 0.2 for Anthropic, so the lead is narrowing. For context, xAI is paid for by 4.1% of businesses and DeepSeek by 0.3%.
The S-1/A adds customer counts as of September 30, 2026:
More than 1,900 customers with $1 million or more of trailing twelve-month revenue
About 250 customers with $10 million or more
More than 11,900 customers with $100,000 or more
AI adoption by industry and company size
Ramp's September 2026 data shows which businesses pay for any AI tool, which is the market Claude sells into:
Segment | Share paying for AI (Sep 2026) | Aug 2026 |
|---|---|---|
Information | 79.9% | 80.0% |
Finance and insurance | 73.7% | 73.1% |
Manufacturing | 60.9% | 60.3% |
Retail trade | 49.9% | 48.3% |
Construction | 43.2% | 41.2% |
Healthcare | 42.5% | 42.3% |
Accommodation and food | 33.0% | 32.6% |
Large businesses | 65.7% | 65.2% |
Medium-sized businesses | 62.5% | 61.4% |
Small businesses | 50.0% | 49.5% |
Source: Ramp AI Index, September 2026. The median firm paying for AI spends $15.70 per employee per month, and the top 1% spend $9,567.53, so spending is concentrated in a small group of heavy users.
Valuation, funding, and the IPO
Anthropic's last private round was a $65 billion Series H at a $965 billion post-money valuation on May 28, 2026, up from $380 billion in February. On October 26, 2026 it filed its S-1 publicly, and amended it on November 2, applying to list Class A common stock on the Nasdaq Global Select Market under the symbol ANT. There was no prior public market for the shares.
Who uses Claude, and for what?
First Page Sage's October 2026 model shows the mix of conversations shifting away from coding as Claude broadens its audience:
Use case | Share (October 2025) | Share (October 2026) | Change |
|---|---|---|---|
Software development and coding | 39.3% | 33.6% | -5.7 pts |
Writing and editing | 13.4% | 16.1% | +2.7 pts |
Research and analysis | 11.8% | 13.3% | +1.5 pts |
Education and tutoring | 9.6% | 10.9% | +1.3 pts |
Business strategy and operations | 7.2% | 9.0% | +1.8 pts |
Data analysis | 5.9% | 6.4% | +0.5 pts |
Personal advice and wellbeing | 5.1% | 4.8% | -0.3 pts |
Creative writing | 4.0% | 3.1% | -0.9 pts |
Other | 3.7% | 2.8% | -0.9 pts |
Source: First Page Sage, Claude Usage Statistics, November 2026. Coding is still the largest category at a third of conversations.
Who uses Claude? Age and gender
Age group | Share of visitors |
|---|---|
18–24 | 20.05% |
25–34 | 30.55% |
35–44 | 19.5% |
45–54 | 14.68% |
55–64 | 9.64% |
65+ | 5.57% |
Source: Similarweb, claude.ai audience data, September 2026. 56.74% of visitors are male and 43.26% are female. The 25–34 bracket is the largest, at 30.55%. Claude's audience is somewhat older than DeepSeek's and more balanced than Grok's, which are 63.31% and 68.34% male on the same measure.
Where are Claude's users?
Country | Share of claude.ai traffic (Similarweb, Sep 2026) | Share of Claude visitors (First Page Sage, Oct 2026) |
|---|---|---|
United States | 25.09% | 21.1% |
India | 10.23% | 10.0% |
United Kingdom | 4.11% | 5.2% |
France | 3.62% | 3.3% |
Japan | 3.38% | 3.2% |
Brazil | — | 4.3% |
Germany | — | 3.9% |
Canada | — | 3.5% |
South Korea | — | 2.9% |
Australia | — | 2.6% |
Sources: Similarweb, September 2026; First Page Sage, November 2026. The two models agree on the top two markets, with the United States and India accounting for about a third of traffic.
Subscriber retention by plan
First Page Sage measured how many subscribers who joined in autumn 2025 still held a paid plan after 3, 6 and 12 months. The figures are directional, because Anthropic does not publish retention.
Plan | 3 months | 6 months | 1 year |
|---|---|---|---|
Claude Pro | 77% | 68% | 61% |
Claude Max | 85% | 77% | 71% |
Claude Team | 88% | 82% | 74% |
Claude Enterprise | 96% | 93% | 89% |
Source: First Page Sage, November 2026. Pro keeps 61% after a year, ahead of the 56% it estimates for ChatGPT Plus. Enterprise keeps 89%.
What does Claude cost?
Plan | Price | Who it's for |
|---|---|---|
Free | $0 | Everyone |
Pro | $20/mo ($17/mo billed yearly) | Everyday work |
Max | $100/mo | All-day work |
Ultra | $200/mo | Your longest tasks |
Team | $25/seat/mo ($20 billed yearly); premium seats $125 ($100 yearly) | Teams of 2 to 150 |
Enterprise | $20/seat/mo, plus usage billed at API rates (20-seat minimum) | Large companies |
Source: claude.com/pricing, November 2026 (US prices before tax).
The latest Claude models
Anthropic released four models in September 2026: Claude Fable 5.1, Mythos 5.1, Opus 5.5 and Sonnet 5.5. Haiku 5.5 now completes the current lineup on its pricing page. Mythos-class models remain on limited availability rather than general release.
Model | API price per million input / output tokens | Context window |
|---|---|---|
Claude Fable 5.1 | $10 / $50 | 1M tokens |
Claude Opus 5.5 | $4 / $20 | 1M tokens |
Claude Sonnet 5.5 | $2 / $10 | 1M tokens |
Claude Haiku 5.5 | From $0.10 / $0.50 | 1M tokens |
Source: Claude Platform docs and claude.com/pricing, November 2026. All four models have a June 2026 knowledge cutoff. For comparison, Gemini 4 Argon launched at $2 and $10, and DeepSeek V4.1-Flash costs $0.30 and $1.20 at peak, per our Gemini and DeepSeek statistics.
How Claude compares to the competition
On consumer reach, Claude still trails ChatGPT and Gemini. Similarweb's share of generative AI website traffic shows how fast the gap has been closing:
Platform | Oct 2025 | Apr 2026 | Jul 2026 | Aug 2026 | Sep 2026 |
|---|---|---|---|---|---|
ChatGPT | 71.8% | 53.7% | 54.3% | 55.6% | 56.4% |
Google Gemini | 13.6% | 26.7% | 26.6% | 25.6% | 25.0% |
Claude | 2.3% | 8.0% | 9.6% | 9.3% | 9.6% |
DeepSeek | 4.1% | 4.0% | 3.4% | 3.5% | 3.4% |
Grok | 2.8% | 3.2% | 2.5% | 2.5% | 2.3% |
Microsoft Copilot | 2.0% | 2.0% | 1.7% | 1.6% | 1.5% |
Perplexity | 2.5% | 1.5% | 1.0% | 0.9% | 0.9% |
Source: Similarweb, gen AI website traffic share, via Rhino Rank, November 2026. Claude is third, and its share has roughly quadrupled in a year. Statcounter's referral data, which counts clicks chatbots send to other websites, shows a very different picture:
Platform | Aug 2026 | Sep 2026 | Oct 2026 |
|---|---|---|---|
ChatGPT | 79.4% | 80.47% | 51.58% |
Google Gemini | 10.93% | 10.94% | 28.86% |
Perplexity | 4.34% | 5.84% | 15.04% |
Claude | 2.53% | 0.95% | 2.3% |
Microsoft Copilot | 2.77% | 1.78% | 2.18% |
Source: Statcounter Global Stats, August to October 2026. BrightEdge's August data is more extreme: Claude was 0.8% of AI referral traffic, and its referrals fell 15% in the month. Claude's referral share has stayed below 4% since June even as its revenue multiplied, which supports the split this whole post keeps returning to: Claude's growth comes from enterprise contracts and API and developer usage, not from casual chatbot browsing that shows up in referral data. On the metrics that matter to Anthropic's business, such as enterprise spend share and revenue, Claude is winning. On the metrics that matter to a typical consumer's daily habits, it is still catching up.
For the fuller competitive picture across all major AI platforms, see our ChatGPT statistics, Google AI Overviews statistics, and Perplexity statistics posts.
Sources
9fin — Anthropic S-1 analysis: revenue gains mask looming infrastructure bill, October 27, 2026
Reuters — Anthropic logged net loss of over $80 billion in first half, October 20, 2026
CNBC — Anthropic says annualized revenue climbed to $65 billion in July, August 17, 2026
Reuters — Anthropic revenue run rate tops $65 billion, August 17, 2026
First Page Sage — Top Generative AI Chatbots by Market Share, November 2026
Rhino Rank — AI statistics report, November 2026 (Similarweb and Statcounter data)
Similarweb — Gen AI website traffic share update, September 2026
BrightEdge — ChatGPT referral traffic more than doubles in 2026, September 24, 2026
Statcounter Global Stats — AI Chatbot Market Share, October 2026