For three years, ChatGPT was the AI market. Now it’s just the biggest player in one, and the gap is closing fast.

ChatGPT’s app market share fell from 69% in January 2025 to 45% in January 2026. On the web, the same trend tracks: from 87% of global AI chatbot traffic to around 64-68%. The quasi-monopoly that defined the first era of consumer AI is over.

The biggest winner is Google. Gemini’s market share nearly quadrupled over the past year, from around 5% to over 21% on the web, and from 15% to 25% on mobile apps. Gemini surpassed 750 million monthly active users last quarter and crossed 2 billion monthly web visits for the first time in January. Google’s advantage is structural: Gemini is embedded across Search, Android, Chrome, Gmail, Docs, and Workspace. Twice as many U.S. Android users engage with Gemini through the operating system as through the standalone app. You don’t need to download anything. It’s already there.

Elon Musk’s Grok is the surprise mover, jumping from 1.6% to 15.2% app market share in a year, now ahead of Perplexity. Claude, Anthropic’s model, holds around 2% of overall web traffic but dominates in engagement: 34.7 minutes per daily user in January, the highest of any platform. And on March 2, Claude overtook ChatGPT as the #1 free app on the U.S. App Store for the first time after OpenAI took the Pentagon deal Anthropic morally refused to.

The numbers behind the scenes are just as telling. About 79% of OpenAI enterprise users also pay for Anthropic. That means Claude isn’t stealing customers: it’s becoming a second tool that businesses use alongside ChatGPT, often for different tasks. Anthropic’s revenue hit $14 billion annualized in February 2026, up from $1 billion roughly 13 months earlier, one of the fastest revenue growth trajectories in tech history. Claude Code, its AI coding tool, alone generates over $2.5 billion in annualized revenue.

OpenAI CEO Sam Altman sent a “code red” memo to staff last year instructing teams to focus on personalization, reliability, and image generation, the exact areas where Gemini and Claude had been gaining ground. The internal alarm confirmed what the market data already showed: the first-mover advantage is eroding.

And OpenAI’s response has been to do more of everything. Ads in ChatGPT Search. Commerce features with Shopify. A $30 billion investment from Nvidia. A potential IPO at a $550-600 billion valuation. A Pentagon contract was secured hours after Anthropic was blacklisted. The strategy is scale, speed, and proximity to power.

But there’s a tension in that approach. The more OpenAI adds – plugins, image generation, custom GPTs, shopping, web browsing, file analysis – the more cluttered the product becomes. User experience studies have noted that ChatGPT’s interface now creates “decision fatigue.” Claude and Gemini, by contrast, have kept their interfaces cleaner. Sometimes the competitive advantage isn’t having more features. It’s having fewer — and making them work.

The AI market is consolidating into what looks like a three-player race: OpenAI for consumer scale, Google for ecosystem integration, and Anthropic for enterprise depth. Everyone else — Microsoft’s Copilot (declining despite its OpenAI partnership), Perplexity (growing fast but niche), Grok (surging on distribution through X) — is fighting for the edges.

OpenAI is still the biggest. But “biggest” and “best-positioned” aren’t the same thing — and for the first time, the market is pricing them differently.