We analyzed LLM market share by combining usage-based data and traffic estimates to show how demand for large language models is distributed across AI labs and AI applications. Here are our observations:
- The United States dominates global LLM usage across both traffic and compute, while China stands out for high token usage.
- OpenAI’s ChatGPT remains the leading application, but is steadily losing share to Google’s Gemini, which demonstrates the most balanced and sustained growth across both consumer and API usage.
- Chat-based applications account for nearly all user traffic, while other categories remain niche and fragmented.
LLM market share comparison by country
The United States dominated web traffic across all four months, consistently accounting for 85.5–90.5%. This reflects both the concentration of consumer AI products in English-speaking markets and the US base of most major platforms.
China shows a different usage profile:
- In November 2025, China generated 50.9% of tokens but 7.5% of traffic. This implies heavy API or programmatic usage rather than consumer activity.
- By February 2026, the gap narrowed. Token share dropped to 31.9%, while traffic remained around 8.0%, suggesting more balanced usage or a shift in how AI is consumed.
LLM market share comparison based on applications
ChatGPT remains the clear leader, but competitive pressure is increasing. Its traffic share dropped from 72.5% in October 2025 to 60.5% in February 2026, a 12-point decline over four months.
Gemini captured most of that shift by growing from 13.9% to 23.9%. The increase is most likely to be driven by distribution and steady model improvements.
LLM market share comparison based on AI labs
OpenAI saw the largest shift in the dataset. Its power index rose from 16.9% in November 2025 to 51.8% in February 2026, driven mainly by a recovery in traffic share (from 19.5% to 57.3%).
Anthropic moved in the opposite direction, dropping from 52.7% to 16.8%
Google shows the most balanced growth across all metrics:
- Power index is increased from 14.2% to 27.4%
- Traffic: 11.9% to 16.5%
- Token share: 16.7% to 25.7%
App market share comparison with categories
Chat dominates the entire market, consistently accounting for 88–92% of traffic. Consumer AI usage is still centered on general-purpose conversational interfaces. Within the Chat category, ChatGPT and Gemini accounted for ~84% of total traffic in February 2026.
How we define LLM market share
LLM market share is not measured using a single indicator. For this analysis, we defined it in terms of two complementary dimensions that capture both supply-side and demand-side dynamics in the large language model market.
Model usage
Model usage is measured using token-level data from OpenRouter. The aim is to show how frequently specific AI models are used in real workloads, particularly in:
- Code generation.
- AI-powered solutions embedded in applications.
- Agent-based and retrieval augmented generation systems (RAG).
Model usage is treated as a proxy for developer and application-level demand, especially in the cloud segment and API-driven environments.
Traffic share
We measured traffic share using Similarweb estimates of web and app traffic. Traffic data reflects user adoption rather than model performance or capability and is particularly relevant for:
- AI chatbots and virtual assistants.
- Content creation tools.
- Search and consumer-facing generative AI applications.
Linking applications to AI labs
To estimate lab-level market share, we mapped AI applications to the AI labs whose models they use. This mapping is weighted rather than binary:
- Applications using a single lab’s models are attributed fully to that lab.
- Applications using multiple models are attributed proportionally based on token usage data.
- This approach allows application traffic to be redistributed to AI labs in a way that reflects actual model usage.
The result is a usage-weighted view of market share that links user-facing demand to the underlying AI infrastructure.
LLM market share methodology
Step 1: Identifying AI labs and AI applications
- AI labs: Organizations that develop and maintain LLMs. For example, OpenAI, Google, Anthropic, DeepSeek, Qwen, X, Mistral AI, and Meta Llama are among the most prominent AI labs developing AI apps.
- AI applications: End-user tools, platforms, or agents that rely on one or more LLMs, such as ChatGPT, Claude, Grok, and Gemini.
Each AI application is then mapped to:
- One or more underlying AI labs.
- One functional category.
Step 2: Mapping AI apps to AI labs
We matched each AI application to the AI labs it relies on. Attribution is weighted, not binary. For example:
- ChatGPT is 100% OpenAI.
- Cursor consists of multiple labs (e.g., OpenAI, Anthropic, others).
Weights are calculated using:
- OpenRouter token usage data
- Relative model usage within each application
This process allowed us to redistribute app-level traffic back to labs and estimate the effective market share at the lab level.
Step 3: Calculating lab-level market share
We pulled traffic data for the apps from Similarweb, distributed it to the labs based on model distribution, and collected the models’ total token usage.
An example calculation:
Assume an app includes a model that is 30% owned by X Lab and 70% by Y Lab. We pull the traffic data from Similarweb and find that the app has 1 million visits. We then assign 300k of that traffic to X Lab and 700k to Y Lab based on the model distribution.
Next, we collect the total number of tokens used for each lab. For instance, X Lab’s models total 3 million tokens, while Y Lab’s models total 5 million tokens. The calculation works as follows:
- X Lab: 300,000 × 3,000,000
- Y Lab: 700,000 × 5,000,000
Afterward, we validate the results to compute the power index.
Step 4: Categorizing AI applications
We grouped AI applications into categories based on their primary use case. While many tools span multiple functions, we assigned each application a single primary category to ensure consistency in data analysis.
- General-purpose chat: Applications focused on conversational interaction, reasoning, and broad task support. These tools account for a large share of consumer-facing LLM usage and play a central role in customer interactions.
- Programming and coding assistants: Tools primarily used for code generation, debugging, and software development workflows. This category is closely linked to enterprise usage and developer productivity.
- Developer platforms and tooling: Platforms that enable developers and enterprises to build, deploy, or manage AI applications. These tools are central to AI integration and are often used by cloud providers and large enterprises.
- Search and answer engines: Applications optimized for information retrieval and synthesis rather than open-ended chat. These tools often combine language models with internet data to achieve higher accuracy.
- Vision and multimodal generation: Applications focused on image and video generation or understanding, often used in content creation and media-related industry verticals.
- Audio and speech: Tools centered on voice generation, speech interaction, or audio processing.
- Gaming and interactive AI: Applications designed primarily for entertainment, role-play, or interactive experiences.
Limitations of LLM market share analysis
This analysis is based on observable usage signals and therefore has several limitations.
- Traffic estimates from Similarweb can be directional and may underrepresent private, enterprise-only, or regionally restricted usage.
- OpenRouter data reflects a subset of global LLM usage and does not capture all enterprise deployments or proprietary integrations.
- Mapping applications to AI labs assumes that token usage shares are representative over the forecast period, even though model preferences can change rapidly.
- Regional markets, such as China and parts of the Asia-Pacific, may be underrepresented due to data availability constraints.
As a result, the figures should be interpreted as relative indicators rather than precise measurements of market size or revenue.
Be the first to comment
Your email address will not be published. All fields are required.