of developers use AI coding assistants or coding agents at work (Stack Overflow 2026 survey)
Sample: 30,903 valid responses from 169 countries, fielded 23 Jun to 5 Aug 2026; self-selected online survey. Each question has its own n.
Developer jobs and pay, tool use, and how fast websites load, from official data and large surveys.
of developers use AI coding assistants or coding agents at work (Stack Overflow 2026 survey)
| Group | Use AI coding assistants or agents |
|---|---|
| 1 to 5 years | 57.8% |
| 6 to 15 years | 72.7% |
| 16+ years | 69.6% |
| All respondents | 65.9% |
Source: Stack Overflow, Stack Overflow Developer Survey, Jun to Aug 2026
of developers use AI coding assistants or coding agents at work (Stack Overflow 2026 survey)
Chart studio
Switch the chart type and the style; sort it, download it or embed it. Hover a bar for the exact value.
| Group | Use AI coding assistants or agents |
|---|---|
| 1 to 5 years | 57.8% |
| 6 to 15 years | 72.7% |
| 16+ years | 69.6% |
| All respondents | 65.9% |
Same data, any style
| Group | Share of agent users |
|---|---|
| Claude Code | 65.5% |
| GitHub Copilot | 58.7% |
| OpenAI Codex | 29.5% |
| Cursor | 25.5% |
| Google Antigravity | 16% |
| Gemini Code Assist | 14.6% |
| Group | Median pay |
|---|---|
| Software developers | 135980$ |
| Web and digital interface designers | 104000$ |
| Software QA analysts and testers | 104300$ |
| Web developers | 92650$ |
| Group | Mobile | Desktop |
|---|---|---|
| 2021 | 32% | 41% |
| 2022 | 31% | 44% |
| 2023 | 36% | 48% |
| Method change: INP replaced FID (Mar 2024) | ||
| 2024 | 44% | 55% |
| 2025 | 48% | 56% |
Every stat
Each card links to its own page with the source, a ready citation and the data.
of developers use AI coding assistants or coding agents at work (Stack Overflow 2026 survey)
Sample: 30,903 valid responses from 169 countries, fielded 23 Jun to 5 Aug 2026; self-selected online survey. Each question has its own n.
of developers with 1 to 5 years of work experience use AI coding assistants or agents, the lowest of any experience group
Sample: n = 4,094 respondents with 1 to 5 years of work experience.
of developers who use AI coding assistants or agents use them every day
Sample: 30,903 valid responses from 169 countries, fielded 23 Jun to 5 Aug 2026; self-selected online survey. Each question has its own n.
of developers who used a coding agent or assistant in the past year used Claude Code; 58.7% used GitHub Copilot
Sample: 30,903 valid responses from 169 countries, fielded 23 Jun to 5 Aug 2026; self-selected online survey. Each question has its own n.
of developers say they trust AI output when they can easily verify it; 6.6% trust it for many tasks including important work decisions
Sample: 30,903 valid responses from 169 countries, fielded 23 Jun to 5 Aug 2026; self-selected online survey. Each question has its own n.
of developers used or planned to use AI tools in their development process in 2025, up from 76% in 2024
Sample: 49,000+ responses from 177 countries; self-selected online survey.
of developers distrusted the accuracy of AI tools in 2025, more than the 33% who trusted it
Sample: 49,000+ responses from 177 countries; self-selected online survey.
faster: freelance developers with GitHub Copilot finished one JavaScript server task in 71 minutes vs 161 without it (2022 test)
Sample: 95 professional freelance developers recruited on Upwork (45 with Copilot, 50 without); the paper says thirty-five from both groups completed the task and survey. Most aged 25-34, mainly from India and Pakistan.
longer: experienced open-source developers took 19% more time on tasks when allowed AI tools (early-2025 test; METR now calls it out of date)
Sample: 16 experienced developers from large open-source repositories, 246 real issues randomly assigned to AI allowed or not.
average security pass rate of AI-written code across more than 100 models tested since 2023, so about 4 in 10 tasks fail (vendor test)
median yearly pay of US software developers, QA analysts and testers in 2025 ($135,980 for software developers alone)
jobs for US software developers, QA analysts and testers in 2025, projected to grow 10% to 2035
median yearly pay of US web developers and digital designers in 2025; web developers alone earned $92,650
of mobile websites had good Core Web Vitals in 2025, against 56% on desktop
Sample: Chrome UX Report real-user data for millions of websites (origins), mobile and desktop.
of mobile pages had good loading speed (Largest Contentful Paint) in June 2025, the weakest of the three Core Web Vitals
Sample: Chrome UX Report real-user data for millions of websites (origins), mobile and desktop.
of all websites use WordPress, a 58.3% share of sites with a known content management system
of websites with a known server-side language use PHP; JavaScript is second at 8.1%
of all websites now run on Lovable, an AI app builder used for vibe coding, slightly more than Astro or GitHub Pages (0.2% each)
developers on GitHub in 2025, after more than 36 million joined in a year (over one per second)
of new developers on GitHub use Copilot in their first week; TypeScript became GitHub's most used language in Aug 2025
Reconciled
The different numbers you see quoted are often all real: they come from different panels, years and questions. Here they are side by side.
| Number | When | Who measured | Why it differs |
|---|---|---|---|
| 55.8% faster | 2022 test, published 2023 | GitHub Copilot experiment (arXiv 2302.06590) | 95 Upwork freelancers, one new JavaScript task, early Copilot. |
| 19% slower | Feb to Jun 2025 | METR randomized trial | 16 experienced developers on their own large repositories, early-2025 tools; METR now calls it out of date. |
| -18% for returning developers | Aug 2025 to early 2026 | METR update, Feb 2026 | Late-2025 tools; METR reads it as some evidence of speed-up but says selection effects make it weak evidence. |
| 73% daily use | Jun to Aug 2026 | Stack Overflow Developer Survey 2026 | Self-reported use, not measured speed. |
| Number | When | Who measured | Why it differs |
|---|---|---|---|
| 33% trust, 46% distrust | 2025 | Stack Overflow Developer Survey 2025 | Five-point trust scale on AI accuracy. |
| 48% trust when easy to verify; 6.6% for important decisions | 2026 | Stack Overflow Developer Survey 2026 | New answer options about when people trust AI; not comparable as a trend. |
Stat check
Famous stats with no real source, or quoted wrong, and where they came from.
GitHub Copilot makes developers 55% more productive
A 2022 controlled test (arXiv 2302.06590): 95 Upwork freelancers recruited, one task (an HTTP server in JavaScript); the Copilot group finished 55.8% faster (95% CI 21% to 89%). It measured one task's speed, not overall productivity. Where it came from
Quote this instead: 55.8%AI makes experienced developers 19% slower
METR's early-2025 trial of 16 open-source developers. METR now labels it out of date; its February 2026 update points to a likely speed-up but says the new data is unreliable because developers refused to work without AI. Where it came from
Quote this instead: 19%Only 29% of developers trust AI tools
Repeated by roundup sites for the 2025 Stack Overflow survey. The survey page says 33% trust AI accuracy and 46% distrust it; in 2026 the question changed and 48% trust AI output when they can easily verify it. Where it came from
Quote this instead: 48%53% of mobile visitors leave a site that takes longer than 3 seconds to load
Google/DoubleClick 'The Need for Mobile Speed' (2016). The Think with Google page that hosted it now redirects elsewhere; use current Core Web Vitals data instead. Where it came from
Quote this instead: 48%Every number on this page was checked at its original source. Free to cite with a link to the stat page.
Free alerts from DigitalTreed, straight to your browser.