If you're investing in AI visibility, the first practical question is which engines to actually invest in. The data on generative AI chatbot usage is now solid enough to answer that question with specificity rather than gut feel. This post pulls together the market share picture as of May 2026, drawing on public traffic data, primary-source category reports, and our own measurement at Honeyb. It then translates those numbers into the only thing that matters: where your AI visibility budget should land.
Top generative AI chatbots by market share, May 2026
The table below ranks the major generative AI chatbots by share of estimated US usage. ChatGPT dominates, but the next four entries (Gemini, Copilot, Perplexity, Claude) collectively account for roughly 38% of usage and are growing faster than the leader.
| Rank | Chatbot | Description | Underlying models | Market share | Quarterly user growth |
|---|---|---|---|---|---|
| 1 | ChatGPT (excluding Copilot) | General-purpose AI chatbot | GPT-5.2, GPT-5, GPT-4.1 | 60.6% | +4% |
| 2 | Google Gemini | General-purpose AI assistant | Gemini 3 | 15.1% | +12% |
| 3 | Microsoft Copilot | General-purpose AI assistant | GPT-5.2, GPT-5.1, Claude Haiku 4.5 | 12.5% | +3% |
| 4 | Perplexity | Accuracy-focused AI search engine | Mistral 7B, Llama 2 | 5.4% | +4% |
| 5 | Claude | Business-focused AI assistant | Sonar, GPT-5.2, Sonnet 4.5 | 5.0% | +14% |
| 6 | Grok | General-purpose AI search engine | Grok 4.1 | 0.6% | +4% |
| 7 | DeepSeek | General-purpose AI search engine | DeepSeek V3.2 | 0.2% | +7% |
| 8 | Brave Leo AI | Privacy-focused AI assistant | Qwen 14B, Llama 3.1, Claude Haiku 3.5 | 0.1% | +3% |
| 9 | Komo | Link-surfacing AI search engine | Not disclosed | 0.1% | +2% |
| 10 | Andi | Simplicity-focused AI search engine | Not disclosed | 0.1% | +4% |
Two structural takeaways jump out. First, this is still a one-and-a-half horse race: ChatGPT plus Copilot (which is ChatGPT underneath, with personalisation) accounts for roughly 73% of usage. Second, the slope of the curve matters more than the rank order. Claude and Gemini are growing at three to four times the rate of the leader, which means the share ranking 18 months from now will not look like this.
A quick note on the chart numbers. The historical trend charts below bundle Microsoft Copilot in with ChatGPT, since Copilot runs on the same underlying ChatGPT models. That's why the ChatGPT line in the next chart hovers around 73% rather than the 60.6% standalone figure in the ranking table. 60.6% (ChatGPT) plus 12.5% (Copilot) is roughly the 73% you'll see on the trend curve.
Market share (%)
Top four generative AI chatbots compared (ChatGPT figure includes Copilot)
The fastest-growing generative AI chatbots
Re-ranking by quarter-over-quarter user growth instead of absolute share makes the trajectory clearer. Claude leads on growth, Gemini is the only top-three engine still expanding meaningfully, and ChatGPT's growth has settled into the low single digits as the category matures around it.
| Rank | Chatbot | Market share | Quarterly user growth |
|---|---|---|---|
| 1 | Claude | 5.0% | +14% |
| 2 | Google Gemini | 15.1% | +12% |
| 3 | DeepSeek | 0.2% | +7% |
| 4 | ChatGPT (excluding Copilot) | 60.6% | +4% |
| 5 | Microsoft Copilot | 12.5% | +3% |
| 6 | Perplexity | 5.4% | +4% |
| 7 | Grok | 0.6% | +4% |
| 8 | Andi | 0.1% | +4% |
| 9 | Brave Leo AI | 0.1% | +3% |
| 10 | Komo | 0.1% | +2% |
The pattern is consistent with what other category data shows. The headline number people remember (ChatGPT has X% share) understates the importance of the second tier. Investing only in ChatGPT visibility right now optimises for the present, not the next two years.
ChatGPT market share (with Copilot bundled in), 2024 to date
Bundled together, ChatGPT plus Copilot has compressed by roughly three points over 28 months, from around 76% in early 2024 to 73% in spring 2026. The decline is gradual but consistent, which is what category leaders look like when smaller specialised competitors take share without ever displacing them. The two are reported together here because Copilot runs on the same underlying ChatGPT models; the only meaningful difference is that Microsoft Copilot is tuned to data inside the Microsoft 365 ecosystem.
Market share (%)
ChatGPT market share (including Microsoft Copilot)
Google Gemini market share, 2024 to date
Gemini's curve is the most interesting one on this chart. After a roughly two-year plateau in the mid-13s through 2024 and most of 2025, Gemini jumped above 14.5% in November 2025 and has stayed around 15% since. The jump tracks Google's integration of AI Mode into core Search and the launch of Gemini 3. If the trajectory holds, Gemini will close the gap with ChatGPT faster than any other engine.
Market share (%)
Google Gemini market share
Perplexity market share, 2024 to date
Perplexity is the most volatile engine in the top tier. It climbed from under 3% in early 2024 to over 6.5% through 2025, then gave back roughly a point and a half in late 2025 and early 2026 as ChatGPT Search and Google AI Mode ate into its core differentiation. It's still net up over 28 months, but the trajectory is sideways now. The interesting question is whether its citation-first design (which AI visibility teams love) survives in a market increasingly dominated by general-purpose assistants with shopping and code generation built in.
Market share (%)
Perplexity market share
Claude market share, 2024 to date
Claude is the cleanest growth story on the chart. From around 2% in early 2024 to 5% by spring 2026, with an inflection in late 2025 tied to Sonnet 4.5 availability and Anthropic's enterprise push. Claude is growing faster than any other top-five engine and its user base skews heavily toward B2B and developer audiences, which makes it disproportionately important for SaaS, infrastructure, and professional services brands.
Want to see this in action?
Check how AI models talk about your brand — free, instant, no signup required.
Market share (%)
Claude market share
What this means for AI visibility prioritisation
Numbers without action are just numbers. Here's how the share distribution translates into where AI visibility budget should actually land in 2026, by audience type.
If you sell to consumers
Prioritise in this order: ChatGPT, Gemini, then Perplexity. ChatGPT is the surface most consumers reach for; Gemini's integration with Google Search means it's the engine non-AI-native consumers encounter without realising. Perplexity matters as a tiebreaker for product research where the user is comparing options. Claude is a lower priority unless your product has a tech-literate or professional audience.
If you sell to businesses (SaaS, B2B, professional services)
Reverse the order at the top. Claude has 5% of total share but a vastly higher proportion of professional, technical, and enterprise users. ChatGPT is still the default, but a brand that appears in Claude and not ChatGPT is invisible to the wrong half of the buying committee. Order: ChatGPT, Claude, Gemini, Perplexity, Copilot. Copilot matters most if your buyers are inside the Microsoft 365 ecosystem.
If you sell to developers
Claude is the leading engine for code generation and developer adjacent tasks, in part because it powers Cursor, Windsurf, and most of the popular agentic coding tools. ChatGPT remains the default for general queries. Perplexity matters for technical documentation discovery. Order: Claude, ChatGPT, Perplexity, then Gemini and Copilot at lower priority.
If you sell e-commerce or physical products
ChatGPT Shopping is the single highest-leverage surface in this group, because the buying journey now starts and often ends inside a single ChatGPT conversation. Perplexity's shopping integration matters second. Gemini matters for product research that originates from Google. Claude is a lower priority since it doesn't yet have shopping surfaces. Order: ChatGPT, Perplexity, Gemini, then Claude.
Where ChatGPT-only strategies break
The simplest critique of how most brands approach AI visibility today: they optimise for ChatGPT and assume the others will follow. They don't. SE Ranking's research found that 89% of citations come from different domains depending on whether you ask ChatGPT or Perplexity. A brand can dominate one engine and be effectively invisible on another. Honeyb's own dataset across 21 buyer-question categories showed only 5 of 21 had a clean #1 across all four major models; the rest showed material disagreement. The share data above reinforces the strategic implication: you cannot pick one engine and call it a strategy.
Methodology and sources
The share figures above aggregate from public AI traffic and usage research. Primary sources used:
- Statista, Artificial Intelligence Market Forecast (2024 and 2026 updates)
- Grand View Research, Artificial Intelligence Market Size, Share & Trends Analysis Report
- Fortune Business Insights, Artificial Intelligence Market Size, Share & Growth Report
- McKinsey, The state of AI in 2023: Generative AI's breakout year
- Artificial Intelligence Market Size and Share Analysis (2024)
- 2024 AI Business Predictions (PwC)
- Artificial Intelligence Market, Size, Global Forecast 2024-2030
- AI Industry Analysis: 50 Most Visited AI Tools and Their Traffic Behavior
Cross-referenced with Similarweb traffic data and First Page Sage's quarterly AI chatbot market share series. The trend charts above pull data points published month-by-month across the cited research. Where sources differed by more than one point, we used the conservative figure.
AI usage shifts quickly, often weekly. The numbers above are accurate as of May 2026. The implication for which engines to invest in usually changes slower than the share numbers themselves, because the underlying audience mix per engine is more stable than the engine share.
Closing
Treat the share table as input, not output. The actionable question is: across the engines that matter for your audience, where do you currently appear, where do you not, and which competitors hold the slots when you're absent. The free AI visibility check runs that diagnostic across the major engines in around 30 seconds. For the broader picture on how AI engines decide which brands to name, see how AI models choose which brands to recommend and what AI actually recommends across 21 buyer questions.
