Meta employees used over 60 trillion tokens in 30 days, with one user alone consumed 280 billion.
Rohan Paul Twitter · Rohan Paul (@rohanpaul_ai) · 2026-07-01
SemiAnalysis data shows Meta employees consumed over 60 trillion tokens in 30 days—one user alone reaching 280 billion—as coding now accounts for more than 70% of OpenAI and Anthropic revenue, signaling a shift toward high-intensity AI power users over mass adoption.
Appears in
Extraction
Topics: agentic-aienterprise-ai-consumptioncoding-toolsai-economics
Claims
- Meta employees collectively consumed over 60 trillion tokens in a single 30-day period.
- One Meta employee alone consumed 280 billion tokens, equivalent to roughly $50,000 per year in token costs at average rates.
- Coding now explains over 70% of OpenAI and Anthropic annual recurring revenue according to SemiAnalysis.
- Most companies set monthly token caps ranging from $250 to $4,000, with power users exhausting limits within days.
- Agentic AI enables a single skilled operator to generate machine work streams that rival an entire small team's compute footprint.
Key quotes
There are real engineers at Meta and other companies spending around $10 million a year per engineer on these coding tools.
Agentic AI changes that math because one skilled operator can create a stream of machine work that runs all day.
Even internally inside our own company, there are some people who have set up these kinds of multi-agent hierarchies and agent loops that look like their own software architecture.