2025: The State of Generative AI in the Enterprise
Menlo Ventures' 2025: The State of Generative AI in the Enterprise (December 2025; n=495 U.S. enterprise AI decision-makers plus a market model) found that companies spent $37 billion on generative AI in 2025, up from $11.5 billion in 2024.
Key findings
- 01Companies spent an estimated $37 billion on generative AI in 2025, up from $11.5 billion in 2024.The figure comes from Menlo's market model built on its survey, not from audited company filings.
- 02The application layer took $19 billion and the infrastructure layer $18 billion.Applications are the user-facing software built on models; infrastructure includes model APIs, training and supporting tools.
- 03Anthropic earned 40% of enterprise LLM spend, OpenAI 27% and Google 21%.OpenAI's share fell from 50% in 2023; Anthropic's rose from 24% in 2024. Menlo is an investor in Anthropic.
- 04Coding was the largest departmental use at $4.0 billion, 55% of departmental AI spend.Coding tools are the clearest example of a use case where companies are paying at scale.
- 0576% of AI use cases were purchased rather than built internally, and 47% of AI deals went to production versus 25% for traditional SaaS.Menlo reads this as a shift from building in-house toward buying finished tools.
By the numbers
$37Bestimated 2025 enterprise generative AI spend
76%of AI use cases purchased rather than built
$4.0Bspent on AI coding tools
What it means for you Draft
For executives at $10–100M companies
Most enterprise AI money is now going to purchased tools rather than in-house builds, which suggests a $10–100M company does not need its own AI engineering team to get started. Coding tools are where spend is most concentrated, so if you employ developers, that is likely the first budget line to examine. Treat the vendor share figures with care: they come from an investor in one of the vendors.
For practitioners
If you are choosing between building and buying, the market has moved strongly toward buying, so a build decision should rest on a clear reason. Model provider share shifted quickly over two years, so avoid locking workflows to a single model where you can.
Limitations
Medium trust.Disclosed sample and fieldwork, but spend totals are modeled and Menlo invests in AI vendors, including Anthropic.
Spend totals are modeled estimates from a survey of 495 large U.S. enterprises, not measured transactions. Menlo invests in Anthropic and other AI companies, and the sample says little about firms under $100M revenue.
About the publisher: Venture capital firm that invests in AI companies, including Anthropic.
Menlo Ventures. "2025: The State of Generative AI in the Enterprise." December 9, 2025. https://menlovc.com/perspective/2025-the-state-of-generative-ai-in-the-enterprise/
Related reports
Technology Report 2026Leaders expect a 95% developer productivity uplift in 1–2 years; current gains are 20–27%. Bain says AI moves bottlenecks into review and coordination.
Ramp AI Index, September 2026: Cracks in the AI Thesis, Part 2Business AI adoption keeps rising but more slowly; heavy spenders cut per-employee AI spend as token prices fell 41% from their 2026 peak.
Gartner Survey Finds Only 22% of Organizations Have Successfully Scaled AI Across Multiple Business Units (press release)85% of functional leaders plan to increase AI spending in 2026 after allocating an average 12% of functional budgets to AI in 2025.