This report draws on research our team conducted between August 4 and September 26, 2026, in which we surveyed 2,350 professional software developers in North America and Europe about the AI coding tools they rely on. We then weighed those responses against publicly disclosed seat counts and usage figures to produce a normalized model of the market.
We define market share here as the percentage of professional developers who name a given tool as their primary AI coding assistant, meaning the tool they use most often for writing, editing, or reviewing production code. Many developers use two or three tools in parallel, so primary-tool share is a stricter measure than adoption, and it tracks the competition between vendors more closely.
The category has reordered itself in under a year. Terminal-based coding agents, which barely registered in mid-2025, now account for the largest block of developer usage, and the incumbent IDE extension model has ceded ground to them. In the sections below, we break down current market share, its trajectory over five quarters, and how the market divides by company size, interface type, and underlying model.
AI Coding Assistant Market Share
In the table below, we break down the market share of the leading AI coding assistants as of October 2026, along with each tool’s change in share since January 2026.
The AI Coding Assistant Market Share, 2026
| AI Coding Assistant | Developer | Primary Interface | Market Share (Oct 2026) | Change vs. Jan 2026 |
| Claude Code | Anthropic | Terminal Agent | 29.4% | +10.9 pts |
| GitHub Copilot | Microsoft / GitHub | IDE Extension | 22.7% | -7.8 pts |
| Cursor | Anysphere | AI-Native IDE | 13.1% | -4.9 pts |
| OpenAI Codex | OpenAI | Terminal & Cloud Agent | 11.8% | +7.6 pts |
| Google Antigravity & Gemini Code Assist | AI-Native IDE | 5.6% | +1.1 pts | |
| JetBrains AI (Junie) | JetBrains | IDE Extension | 4.3% | -0.6 pts |
| Windsurf | Cognition | AI-Native IDE | 2.9% | -1.9 pts |
| Amazon Q Developer & Kiro | Amazon | IDE Extension | 2.6% | -0.8 pts |
| OpenCode | Open Source | Terminal Agent | 2.4% | +1.6 pts |
| Cline | Open Source | IDE Extension | 1.7% | -0.5 pts |
| All Other Tools | Various | Various | 3.5% | -4.7 pts |
- We found that Claude Code overtook GitHub Copilot as the most common primary AI coding assistant in mid-2026 and has widened its lead since.
- Our data showed OpenAI Codex posting the second-largest gain of any tool, nearly tripling its share since January as developers adopted its cloud task mode.
- We observed the four largest tools holding 77.0% of the market combined, up from 71.2% in January, a sign that the category is consolidating.
AI Coding Assistant Market Share, Q3 2025 to Q3 2026
In the table below, we track the quarterly market share of the five leading tools over the past year. The line graph beneath it plots the same series.
The AI Coding Assistant Market Share, Q3 2025 to Q3 2026
| AI Coding Assistant | Q3 2025 | Q4 2025 | Q1 2026 | Q2 2026 | Q3 2026 |
| GitHub Copilot | 36.2% | 33.4% | 30.1% | 26.0% | 23.1% |
| Claude Code | 7.9% | 12.6% | 19.2% | 24.8% | 28.7% |
| Cursor | 19.4% | 19.9% | 17.6% | 15.2% | 13.5% |
| OpenAI Codex | 1.8% | 3.1% | 4.6% | 8.3% | 11.2% |
| 3.6% | 4.1% | 4.0% | 4.9% | 5.4% |
Quarterly figures are averaged across each quarter, so the Q3 2026 values differ slightly from the October snapshot above. The crossover between GitHub Copilot and Claude Code falls between Q2 and Q3 2026 in quarterly terms. Cursor’s share peaked in Q4 2025 and has declined in each quarter since, while OpenAI Codex accelerated sharply in Q2 2026. Google’s line is the steadiest in the set, with modest gains that track the release of its Antigravity IDE.
AI Coding Assistant Market Share by Company Size
In the table below, we break down primary-tool share by the size of the developer’s employer, measured in total employees.
The AI Coding Assistant Market Share by Company Size, 2026
| Company Size | Claude Code | GitHub Copilot | Cursor | OpenAI Codex | All Other Tools |
| Startup (1-50) | 36.8% | 9.7% | 21.4% | 15.2% | 16.9% |
| Small Business (51-500) | 33.1% | 17.5% | 16.2% | 13.4% | 19.8% |
| Mid-Market (501-5,000) | 27.9% | 25.8% | 11.7% | 10.9% | 23.7% |
| Enterprise (5,000+) | 22.4% | 35.6% | 8.3% | 9.1% | 24.6% |
- We found that GitHub Copilot remains the leading tool at enterprises with more than 5,000 employees, where existing Microsoft licensing agreements make it the default procurement choice.
- Our data showed Claude Code’s share falling steadily as company size increased, from 36.8% at startups to 22.4% at enterprises.
- We observed Cursor holding 21.4% of startups but only 8.3% of enterprises, the widest gap between segments of any tool in our dataset.
AI Coding Assistant Market Share by Interface Type
In the table below, we group every tool in our dataset by the interface through which developers use it and compare each interface type’s share in January and October 2026.
The AI Coding Assistant Market Share by Interface Type, 2026
| Interface Type | Share (Jan 2026) | Share (Oct 2026) | Change |
| Terminal / CLI Agent | 21.3% | 38.6% | +17.3 pts |
| IDE Extension | 41.2% | 29.4% | -11.8 pts |
| AI-Native IDE | 24.8% | 17.9% | -6.9 pts |
| Cloud / Background Agent | 5.1% | 9.2% | +4.1 pts |
| Browser-Based App Builder | 7.6% | 4.9% | -2.7 pts |
Tools that span several interfaces were assigned to the interface each respondent reported using most often, so a developer running Codex mainly from the terminal counts toward the terminal category even though Codex also offers a cloud mode. The shift we measured is behavioral as much as it is vendor-driven: developers are handing longer, multi-file tasks to autonomous agents and reviewing the output, rather than accepting line-by-line suggestions inside an editor. That pattern mirrors what we see across the broader AI agent market, where task-completing agents are gaining on assistive tools.
AI Coding Assistant Market Share by Underlying Model
In the table below, we break down AI coding usage by the family of large language model that generated the code, regardless of which tool the developer used to access it.
The AI Coding Assistant Market Share by Underlying Model, 2026
| Model Family | Developer | Share (Jan 2026) | Share (Oct 2026) |
| Claude | Anthropic | 49.2% | 47.3% |
| GPT | OpenAI | 22.4% | 28.6% |
| Gemini | 11.8% | 10.2% | |
| Open-Weight Models (Qwen, DeepSeek, Kimi, GLM) | Various | 10.3% | 9.4% |
| Grok | xAI | 3.4% | 2.1% |
| All Other Models | Various | 2.9% | 2.4% |
- We found that Claude models generated nearly half of all AI-written code in our dataset, because they power Claude Code and are also a popular model choice inside Cursor, GitHub Copilot, and most open-source agents.
- Our data showed OpenAI’s GPT models gaining 6.2 points of model share, closely tracking the growth of Codex.
- We observed open-weight models holding close to a tenth of coding usage, concentrated among cost-sensitive teams and self-hosted deployments; our Claude usage statistics and top AI models reports cover the broader model landscape.
Requesting a Copy of This Report
If you’d like to request a PDF copy of this report or learn more about our agency, you can reach out here.
- First Page Sage Research Study. October 2026. San Francisco, California.
- Mikhail Bogdanov, JetBrains Research. AI Coding Agents: Adoption Trends. August 2026. Prague, Czech Republic.
- Stack Overflow. Stack Overflow Developer Survey 2026. 2026. New York, New York.
- Sofiah Nichole Salivio, IT Brief UK. Stack Overflow survey finds developers wary of AI tools. October 6, 2026. Auckland, New Zealand.
- GitHub. GitHub Copilot. October 2026. San Francisco, California.
- Anthropic. Claude Code. October 2026. San Francisco, California.
