Skip to content
Neural Daily

BusinessNews

Menlo: enterprise AI spending hits $176.8 billion

🇭🇺 Olvasd el magyarul →

Menlo Ventures says enterprise generative AI spending grew 5.4x in a year. Anthropic's share of the LLM API market jumped to 55%.

By: Mikola Máté Patrik · · 2 min read

Menlo Ventures chart of enterprise generative AI spending growth
Image: Menlo Ventures · Press kit · Chart from the 2026 State of Enterprise AI report · Source

What happened?

On November 3, Menlo Ventures released its 2026 State of Enterprise AI report (Menlo Ventures). It estimates enterprise generative AI spending reached $176.8 billion by September 2026, 5.4 times the $33 billion of 2025. A year earlier, growth was 3.9x.

Key takeaways

  • Infrastructure: $99.5 billion; applications: $77.3 billion.
  • Two categories make up 76% of application spending: coding ($34.0 billion) and assistants ($25.0 billion).
  • Agents now generate 80% of OpenRouter tokens; agent traffic grew 37x since February versus 5x for human traffic.
  • Frontier labs capture 69% of spending, up from 50%.
  • Healthcare AI spending stands at $3.7 billion.

Anthropic leads

In the enterprise LLM API market, Anthropic's share rose from 40% to 55% and OpenAI's from 27% to 31%. Google fell from 21% to 9%. Open-source models account for about 19% of usage but only 4% of spending.

Menlo Ventures chart of enterprise LLM API market share by provider
Menlo Ventures · Press kit · Chart from the 2026 State of Enterprise AI report · Source

Enterprise LLM API market share (source: Menlo Ventures).

Where the money comes from

55% of AI spending comes from new budgets and 22% from existing software budgets. The share of paying enterprises rose from 55% to 65%. AI-native companies use AI in 96% of coding work, while at incumbents IT (91%), data science (88%) and HR (68%) lead adoption. Every dollar spent on AI comes with $1.70 in deployment costs, $102.5 billion in total.

The report is based on a survey of about 400 US enterprise decision-makers plus Menlo's own estimates, so the figures are best read as directional.

Why it matters

Money is flowing into coding and agents, and the market is concentrating around two or three providers. Anyone picking a model provider now faces a growing lock-in risk. The low spending share of open models also shows how much self-hosting can save.

    Produced with the help of AI, reviewed by an editor. How we make our articles

    Reviewed and edited by the Neural Daily desk. About

    Related articles