❯About You
Age, experience, location, and background
How many years of experience do you have as a developer?
3–5 years30%
30%
6–10 years19%
19%
1–2 years17%
17%
Less than 1 year13%
13%
11–15 years13%
13%
16–20 years4%
More than 20 years4%
What is your current role?
Individual contributor (developer / engineer)43%
43%
Founder13%
13%
Data / ML / AI engineer11%
11%
Tech lead / staff / principal engineer9%
DevOps / SRE / Platform engineer6%
Product manager6%
Student4%
C-level / VP4%
Other (please specify)4%
❯Languages & Tools
Programming languages, AI tools, models, and MCP
What programming languages do you use?
JavaScript / TypeScript70%
70%
Python60%
60%
Go15%
15%
Java13%
13%
Rust10%
C/C++8%
PHP8%
C#5%
Scala5%
Other5%
Kotlin3%
Dart3%
Which coding tools do you currently use?
Select all that apply
💡 On average, developers use 2.4 tools. The most popular is Claude Code (Agent), followed by Cursor (IDE).
Claude Code (Agent)57%
57%
Cursor (IDE)43%
43%
GitHub Copilot (Extension)30%
30%
Codex (Agent)23%
23%
Antigravity (IDE)14%
14%
Lovable (App Builder)14%
14%
Cline (Extension)11%
11%
Zed (IDE)9%
Amazon Q Developer (Extension)9%
Kiro (IDE)7%
Replit (App Builder)7%
Windsurf (IDE)5%
CodeRabbit (Code Review)5%
Devin (Agent)2%
Augment Code (Extension)2%
Which AI coding tool are you most likely to recommend?
Select one from the list, or specify in "Other tools" below
Claude Code (Agent)43%
43%
Cursor (IDE)36%
36%
GitHub Copilot (Extension)9%
Kiro (IDE)4%
Codex (Agent)2%
Devin (Agent)2%
Augment Code (Extension)2%
CodeRabbit (Code Review)2%
Which capabilities would you most like to see improved in AI coding agents?
Select all that apply
Fewer hallucinations49%
49%
More affordable pricing43%
43%
Better understanding of large codebases41%
41%
Better memory of prior work32%
32%
More reliable refactoring27%
27%
Safer code execution (sandboxes, permissions)24%
24%
Better collaboration features (team context, shared agents)22%
22%
Better integration with CI/CD and DevOps tools19%
19%
Reasoning16%
16%
Better integration with my IDE11%
11%
Other (please specify)5%
Which models do you use?
Select all that apply
Opus 4.561%
61%
Sonnet 4.551%
51%
Gemini 3 Pro37%
37%
GPT-5.134%
34%
GPT-5.1 Codex High24%
24%
Composer 115%
15%
GPT-5.1 Codex Mini7%
Grok Code7%
Kimi K25%
I don't know which model my tool uses2%
❯Vibe Engineering
How you use AI coding tools
How often do you use AI coding tools?
Yes, daily87%
87%
Yes, multiple times per week11%
11%
I've tried them but don't use them anymore2%
How much time do you spend reviewing AI-generated code?
1–3 minutes per response45%
45%
I don't, I fully trust the generated code17%
17%
Less than 1 minute per response13%
13%
3–5 minutes per response13%
13%
More than 5 minutes per response9%
Not sure4%
What are your primary reasons for using AI coding tools or agents?
Select all that apply
To code faster75%
75%
To reduce repetitive tasks56%
56%
To prototype ideas quickly56%
56%
To better understand existing code47%
47%
To generate tests or documentation automatically44%
44%
To learn or improve my skills42%
42%
To improve code quality33%
33%
To explore unfamiliar languages or frameworks22%
22%
📊 Selection Analysis
How many do developers use?
Average: 3.8
1
11%
2
8%
3
31%
4
19%
5
17%
6
6%
7
6%
8
3%
Most common tool combinations
#150%
To code faster
+
To reduce repetitive tasks
#244%
To code faster
+
To prototype ideas quickly
#333%
To code faster
+
To generate tests or documentation automatically
#433%
To generate tests or documentation automatically
+
To prototype ideas quickly
#533%
To code faster
+
To better understand existing code
What are the biggest challenges you face when using AI coding tools?
Select all that apply
Inaccurate or hallucinated code58%
58%
Cost of AI tools42%
42%
Hard to debug AI-generated code39%
39%
Limited understanding of my codebase29%
29%
Security or privacy concerns24%
24%
Corporate restrictions / compliance11%
11%
Poor integration with my environment11%
11%
Lack of trust from my team / reviewers8%
I face no significant challenges5%
Other (please specify)5%
📊 Selection Analysis
How many do developers use?
Average: 2.3
1
34%
2
26%
3
26%
4
5%
5
5%
7
3%
Most common tool combinations
#129%
Inaccurate or hallucinated code
+
Limited understanding of my codebase
#229%
Cost of AI tools
+
Inaccurate or hallucinated code
#324%
Inaccurate or hallucinated code
+
Hard to debug AI-generated code
#421%
Cost of AI tools
+
Hard to debug AI-generated code
#516%
Inaccurate or hallucinated code
+
Security or privacy concerns
Does your company have formal guidelines or policies on the use of AI coding tools?
Yes — AI tools are officially encouraged67%
67%
No — No policy exists17%
17%
Yes — AI tools are allowed with restrictions13%
13%
Yes — AI tools are banned2%
How have AI coding tools or agents affected your productivity?
Significantly increased my productivity65%
65%
Moderately increased my productivity20%
20%
Slightly increased my productivity13%
13%
Slight decrease in productivity2%
AI coding tools or agents make you a better developer:
Strongly agree54%
54%
Agree28%
28%
Neutral13%
13%
Disagree2%
Strongly disagree2%
How do you measure the impact of AI coding tools on your productivity?
Weekly hours saved51%
51%
Human-equivalent hours completed by agents43%
43%
Number of commits / PRs29%
29%
Number of tasks assigned to agents17%
17%
Code maintainability17%
17%
Change failure percentage14%
14%
AI cost and savings11%
11%
📊 Selection Analysis
How many do developers use?
Average: 1.8
1
57%
2
23%
3
11%
4
3%
5
3%
7
3%
Most common tool combinations
#117%
Weekly hours saved
+
Human-equivalent hours completed by agents
#211%
Number of commits / PRs
+
Number of tasks assigned to agents
#311%
Change failure percentage
+
Number of commits / PRs
#411%
Human-equivalent hours completed by agents
+
Code maintainability
#511%
Change failure percentage
+
Weekly hours saved
Which tasks would you feel comfortable fully delegating to an autonomous coding agent?
Writing boilerplate code49%
49%
Refactoring code with tests46%
46%
Generating documentation41%
41%
Investigating bugs38%
38%
Reviewing pull requests33%
33%
Setting up a project or environment31%
31%
None — I only want assisted, not autonomous agents28%
28%
Creating CI/CD pipelines26%
26%
Running migrations or database tasks18%
18%
Managing deployments15%
15%
Other (please specify)3%
📊 Selection Analysis
How many do developers use?
Average: 3.3
1
33%
2
13%
3
15%
4
15%
5
5%
6
5%
7
3%
8
5%
9
5%
Most common tool combinations
#133%
Writing boilerplate code
+
Refactoring code with tests
#231%
Generating documentation
+
Refactoring code with tests
#331%
Investigating bugs
+
Refactoring code with tests
#428%
Writing boilerplate code
+
Generating documentation
#528%
Writing boilerplate code
+
Investigating bugs
🔗Language → Tool Preferences
Which AI coding tools do developers prefer based on their programming language?
JavaScript / TypeScript
Top AI tools used:
1.Claude Code (Agent)
61%
2.Cursor (IDE)
50%
3.GitHub Copilot (Extension)
32%
4.Codex (Agent)
25%
5.Cline (Extension)
18%
Python
Top AI tools used:
1.Claude Code (Agent)
67%
2.Cursor (IDE)
42%
3.Codex (Agent)
29%
4.GitHub Copilot (Extension)
25%
5.Cline (Extension)
17%
Go
Top AI tools used:
1.Claude Code (Agent)
67%
2.Cursor (IDE)
33%
3.Zed (IDE)
33%
4.GitHub Copilot (Extension)
33%
5.Amazon Q Developer (Extension)
33%
Java
Top AI tools used:
1.GitHub Copilot (Extension)
60%
2.Claude Code (Agent)
60%
3.Codex (Agent)
40%
4.Kiro (IDE)
20%
5.Zed (IDE)
20%