Home/Insights/AI Market Share Tracker
Insights — Updated Quarterly

AI Market Share Tracker

Enterprise LLM spend and consumer AI chatbot traffic are moving in opposite directions. We track both, every quarter, so you can see where the market actually is — not where the headlines say it is.

Chris Metcalf
By Chris Metcalf · LinkedIn Founder & Director, Aston Reach
Last updated: 7 September 2026
Next refresh: November 2026

Key facts, at a glance

  • Anthropic's Claude leads enterprise LLM spend with 40% share in Q1 2026, up from around 16% two years earlier — overtaking OpenAI (27%).
  • Google's Gemini holds 21% of enterprise LLM spend and 27.9% of consumer chatbot traffic.
  • ChatGPT (OpenAI) still leads consumer AI chatbot traffic with 53.9% share as of May 2026 — down from roughly 87% in early 2025.
  • The company winning enterprise budget is no longer the company winning consumer mindshare.

Most of the public conversation about "who's winning" in AI treats consumer chatbot usage as a proxy for the whole market. It isn't. The provider enterprises are actually paying to run their AI workloads on has moved in a different direction entirely — and if you're making a vendor decision based on which chatbot gets the most headlines, you're looking at the wrong number.

Enterprise

Enterprise LLM spend share

Enterprise budget tells a different story than the consumer headlines. Anthropic's Claude has grown from roughly 16% of enterprise LLM spend two years ago to 40% in Q1 2026 — overtaking OpenAI, which has fallen to 27%. Google's Gemini holds 21% of enterprise spend.

Enterprise LLM spend share — Q1 2026
ProviderSpend share
Claude (Anthropic)40%
ChatGPT (OpenAI)27%
Gemini (Google)21%
Others12%

Source: Menlo Ventures 2026 State of Enterprise AI; Ramp AI Index, Q1 2026.

Where's Microsoft?

Microsoft doesn’t show up in either consumer or enterprise chart, not because it’s small, but because both measurement methods are structurally blind to how Copilot is built, used and how it’s measured.

On the enterprise side, spend indices like Ramp's work from actual corporate card and expense transactions. Microsoft 365 Copilot is typically billed as part of a general Microsoft 365 subscription line, not tagged as a discrete AI expense — so it's largely invisible to a survey built to isolate "AI spend." And when a business calls OpenAI's models via Azure, that spend gets attributed to OpenAI, not Microsoft, since it's OpenAI's model doing the work. Copilot's real enterprise revenue is mostly seat-based licensing, which this kind of index isn't designed to see at all — not because it's a small business, but because it isn't structured as visible per-model API spend.

Consumer

Consumer AI chatbot traffic share

Consumer mindshare tells the opposite story. ChatGPT remains the default consumer AI chatbot by a wide margin, but its dominance has eroded sharply — from an estimated 87% of consumer chatbot traffic in early 2025 down to 53.9% as of the latest reading, with Google's Gemini the clearest beneficiary, now at 27.9%.

Consumer AI chatbot traffic share — as of May 2026
ProviderTraffic share
ChatGPT (OpenAI)53.9%
Gemini (Google)27.9%
Claude (Anthropic)9.2%
Others9.0%

Source: Similarweb, consumer chatbot web traffic, May 2026 reading.

Where's Microsoft?

Microsoft doesn’t show up in either consumer or enterprise chart, not because it’s small, but because both measurement methods are structurally blind to how Copilot is built, used and how it’s measured.

Similarweb does track Microsoft Copilot as its own line when it publishes granular data — but it shows up at around 1.3% of web visits, because that only counts standalone visits to copilot.microsoft.com. The overwhelming majority of real Copilot usage happens inside Windows, Edge, Bing and Microsoft 365 apps, which a website-traffic panel simply can't see. So the "Others" slice in this chart isn't a fair read of Copilot's actual reach — it's just whatever thin sliver of its web-visit traffic happens to get picked up.

What This Means

Consumer mindshare and enterprise trust are no longer the same contest

Anthropic wins the procurement conversation. OpenAI still wins the front page. For leadership teams choosing an AI vendor, that split matters more than either headline number on its own — what technical and security teams are actually choosing to build on and pay for at scale is not the same signal as consumer brand familiarity.

This is exactly the kind of decision we help CIOs, CTOs and CFOs work through: not which tool is winning the news cycle, but which platform choice, adoption approach and governance model will actually deliver measurable value in your organisation — tool agnostic, structured, and tied to real outcomes.

Methodology & Sources

How this tracker is built

Enterprise LLM spend share is sourced from the Menlo Ventures annual State of Enterprise AI report and the Ramp AI Index, which tracks actual enterprise card and API spend on LLM providers.

Consumer AI chatbot traffic share is sourced from Similarweb's web traffic estimates for the leading consumer chatbot properties (chatgpt.com, gemini.google.com, claude.ai and others), reflecting relative share of visits.

This page is refreshed quarterly with the latest available reading from each source. Each quarter's snapshot is also archived at a dated URL so you can see how the picture has shifted over time.

Choosing an AI vendor?

Don't let a headline number make your AI platform decision

We help CIOs, CTOs and CFOs align AI vendor choice and adoption to real business requirements — structured, measured, governed, and tool agnostic. Book a free 30-minute AI discovery call.

Explore AI Transformation  → Book a free AI discovery call  →