Key Takeaways
- • Bessemer Venture Partners unveiled the inaugural AI 100, ranking the top 100 private AI companies, with a combined valuation of $3.6 trillion, more than three times the $1.1 trillion 2025 Cloud 100 it succeeds.
- • The average AI 100 company is valued at $36 billion versus $11.2 billion for the 2025 Cloud 100, while the median is $9.3 billion versus $5 billion a year earlier.
- • Anthropic tops the list, followed by OpenAI, Databricks, Stripe, Waymo, Cognition, Anduril, Ramp, Fireworks and ElevenLabs; these 10 companies average $250 billion each and account for 70% of total value.
- • Frontier labs and foundation model builders make up 9% of the list but hold 55% of its value, while the AI 100 carries a record 57x average ARR multiple, against a median of 23x.
- • The average company grew ARR 480% year over year (median 275%), and Bessemer calls the AI 100 cheapest on a growth-adjusted basis, at an average 12x multiple.
Summary
Bessemer Venture Partners, in partnership with Newcomer Media, published the inaugural AI 100 on November 12, 2026, at the Cerebral Valley AI Summit in San Francisco. The list succeeds the Cloud 100, which Bessemer published annually from 2016 to 2025. Companies were scored on market leadership (35%), estimated valuation (30%), operating metrics (20%) and people and culture (15%). The list's aggregate value of $3.6 trillion is entirely private value.
Valuation is highly concentrated. Anthropic, OpenAI, Databricks, Stripe, Waymo, Cognition, Anduril, Ramp, Fireworks and ElevenLabs together account for 70% of total list value. Frontier labs and foundation model builders represent roughly $1,974 billion in aggregate, which Bessemer describes as the most concentrated category in a decade of its benchmarking. By contrast, the application layer is worth $629 billion across roughly 40 companies, at an average valuation of $15.7 billion, above the $9.3 billion list median.
Growth and multiples are the report's other headline metrics. The AI 100 carries an average 57x ARR multiple, a record that exceeds the Cloud 100's 2021 peak of 34x; the median is 23x, against 20x for the 2025 Cloud 100. Bessemer attributes the premiums to growth, with average ARR growth of 480% and median growth of 275%. On a growth-adjusted basis, Bessemer says the AI 100 is the cheapest cohort in recent history, at an average 12x.
Capital intensity is also high. Six companies have each raised more than $10 billion, led by OpenAI at $224 billion and Anthropic at $130 billion; together the two labs account for $354 billion, about 80% of the six's combined total. Excluding them, the average company has raised $3.9 billion and the median is just over $1 billion. Bessemer also reports that 14 of the 100 companies were founded by alumni of other companies on the list, citing OpenAI and Stripe as key incubators.
Bessemer surveyed more than 300 AI-native companies as part of the application process. Respondents said 79% believe frontier model quality will converge within two to three years, 85% expect compute demand to keep outstripping supply, and 87% expect value to accrue to the application layer rather than model providers. Bessemer identifies cybersecurity, biotech and robotics as the biggest AI opportunities ahead, noting that physical AI companies (13 on the list) hold a cumulative valuation of $332 billion, about 9% of list value.
Bessemer frames the AI 100 as a new kind of benchmark, reflecting faster shipping velocity, shorter time-to-scale and heavier capital requirements than prior cloud cohorts. The disclaimer notes that some companies discussed are current or former Bessemer portfolio companies.
Valuation is highly concentrated. Anthropic, OpenAI, Databricks, Stripe, Waymo, Cognition, Anduril, Ramp, Fireworks and ElevenLabs together account for 70% of total list value. Frontier labs and foundation model builders represent roughly $1,974 billion in aggregate, which Bessemer describes as the most concentrated category in a decade of its benchmarking. By contrast, the application layer is worth $629 billion across roughly 40 companies, at an average valuation of $15.7 billion, above the $9.3 billion list median.
Growth and multiples are the report's other headline metrics. The AI 100 carries an average 57x ARR multiple, a record that exceeds the Cloud 100's 2021 peak of 34x; the median is 23x, against 20x for the 2025 Cloud 100. Bessemer attributes the premiums to growth, with average ARR growth of 480% and median growth of 275%. On a growth-adjusted basis, Bessemer says the AI 100 is the cheapest cohort in recent history, at an average 12x.
Capital intensity is also high. Six companies have each raised more than $10 billion, led by OpenAI at $224 billion and Anthropic at $130 billion; together the two labs account for $354 billion, about 80% of the six's combined total. Excluding them, the average company has raised $3.9 billion and the median is just over $1 billion. Bessemer also reports that 14 of the 100 companies were founded by alumni of other companies on the list, citing OpenAI and Stripe as key incubators.
Bessemer surveyed more than 300 AI-native companies as part of the application process. Respondents said 79% believe frontier model quality will converge within two to three years, 85% expect compute demand to keep outstripping supply, and 87% expect value to accrue to the application layer rather than model providers. Bessemer identifies cybersecurity, biotech and robotics as the biggest AI opportunities ahead, noting that physical AI companies (13 on the list) hold a cumulative valuation of $332 billion, about 9% of list value.
Bessemer frames the AI 100 as a new kind of benchmark, reflecting faster shipping velocity, shorter time-to-scale and heavier capital requirements than prior cloud cohorts. The disclaimer notes that some companies discussed are current or former Bessemer portfolio companies.
Go deeper than the headline
Research every firm in this story, track the ones that matter and get the daily briefing — free for 7 days.