Live statistics
Tracking the money behind the AI industry — spend, revenue, and the gap between them.
Last data update: 2026-07-02Is AI profitable?
The numbers answer: cumulative spend vs revenue across the frontier AI industry.
$0
AI industry net losses since you opened this page
$1.49T
Total AI spend
$613B
Total AI revenue
$875B
Net gap
1 / 12
Profitable companies
3.6B
Monthly active users
$11
Avg monthly cost / user
Key insights
$2.4
spent for every $1 of revenue
74%
of all AI spend comes from just 4 companies
~$1.6B
the AI industry loses every day
456×
DeepSeek is cheaper per user than Microsoft
3.6B
monthly active users across AI products
1 / 12
only one of twelve is AI-profitable: Nvidia
Sort by
Spend
Revenue
Amazon
Full AI capex est. since 2022
—
-$291B
Cumulative P&L
Lost since you opened this page
Alphabet (Google)
Full AI capex est. since 2022
$14
per user / mo
900M monthly users
-$262B
Cumulative P&L
Lost since you opened this page
Microsoft
Full AI capex est. since 2022
$58
per user / mo
218M monthly users
-$235B
Cumulative P&L
Lost since you opened this page
Meta
Full AI capex est. since 2022
$9.9
per user / mo
1.2B monthly users
-$227B
Cumulative P&L
Lost since you opened this page
Oracle
Cumulative est. since 2023
—
-$39B
Cumulative P&L
Lost since you opened this page
OpenAI
Cumulative est. since 2020
$1.2
per user / mo
1B monthly users
-$27B
Cumulative P&L
Lost since you opened this page
Anthropic
Cumulative est. since 2021
$24
per user / mo
30M monthly users
-$27B
Cumulative P&L
Lost since you opened this page
xAI
Cumulative est. since 2023
$8.6
per user / mo
117M monthly users
-$19B
Cumulative P&L
Lost since you opened this page
Mistral AI
Cumulative est. since 2023
$1.7
per user / mo
15M monthly users
-$600M
Cumulative P&L
Lost since you opened this page
Cohere AI
Cumulative est. since 2020
—
-$300M
Cumulative P&L
Lost since you opened this page
DeepSeek
Cumulative est. since 2023
$0.13
per user / mo
130M monthly users
-$200M
Cumulative P&L
Lost since you opened this page
Nvidia
Cumulative est. since 2023
—
+$253B
Cumulative P&L
Earned since you opened this page
Methodology & caveats
- Spend includes direct R&D, compute, and capital expenditure on AI infrastructure (data centres, chips, and networking). Capex is counted as spend despite its long-term asset value — intentionally, to show the scale of capital committed before returns materialise.
- The AI economy is circular: Google funds Anthropic, which runs on Google Cloud; Amazon also backs Anthropic; Microsoft co-invests with OpenAI. Aggregate figures therefore double-count some revenue flows.
- This page tracks whether AI investment specifically has broken even — not company-wide profitability. That is why Amazon and Google appear deep in the red despite being hugely profitable companies overall.
- Figures are industry estimates drawn from sources such as Bloomberg, WSJ, The Information, and Epoch AI, and are refreshed periodically.
Calculator
Count tokens and estimate AI model costs — another handy tool from Awamer.
References & sources
Primary sources by company
Amazon
Alphabet (Google)
Microsoft
Meta
Oracle
OpenAI
Anthropic
xAI
Mistral AI
DeepSeek
Core idea inspired by the site isaiprofitable.com.