AI Developer Tools Benchmark
Make your next product and marketing decision with real developer data
Build your next AI feature or tool on evidence, and avoid the costly mistakes of building blind.
Join the waitlist for our latest developer survey on AI tools, out October 2026.
If you're one of these, you can't miss this
Product and marketing leaders
Choose which AI developer tools to integrate or partner with, backed by data your board accepts.
Developer relations and research
See where your product has to show up in developers' AI workflows.
What is the AI Developer Tools Benchmark?
A standalone survey of professional developers that measures the AI developer tools market as developers live it: which tools they know, which they use, which they would choose, and how far they trust them.
Most teams sizing up this market start with public leaderboards such as SWE-bench or LMArena. Those rank how well a model or agent completes a fixed set of test tasks. They can't tell you how many developers know your tool, how many of those go on to use it, which tool they would pick next, or whether they trust what it writes. Only developers can answer that, so this survey asks them, with the same questions for every tool so the results compare like for like.
- 2,063professional developers who use AI developer tools
- 19AI developer tools, measured on the same questions
- Q3 2026the latest wave, with results out in October 2026
Know what developers actually use
- Only hidden until the results are out% of developers whose employer mandates Copilot, Codex or Claude Code say it is the tool they would pick themselves.
- The top barrier to agent autonomy is hidden until the results are out.
- The most common way developers work with agents is hidden until the results are out.
- The tool with the highest productivity score is hidden until the results are out.
- The tool that turns the most aware developers into users is hidden until the results are out.
Every major AI developer tool, compared
19 tools. Same questions.
One tool converts more than half of aware developers into active users
AI Developer Tools: Conversion Benchmark
% of professional developers using AI developer tools (n=2,063)
UsingAware, not usingConversion
The 19 tools, A to Z
“But we already have leaderboards for this…”
Leaderboards such as SWE-bench, Terminal-Bench and LMArena's WebDev Arena score what an AI model or agent can do on a fixed set of tasks. None of them tells you which tools developers know, use, prefer or trust at work. This benchmark does. The KPIs include:
Adoption. The share of professional developers using each tool.
Awareness. The share of professional developers who know each tool.
Preference. The tool developers prefer to use.
Trust. How far developers trust the output of AI tools.
Answer the questions your board will ask
Which AI tools do we integrate with first?
Use and awareness of every tool.
Do developers actually prefer the tool they use?
Use against preference, tool by tool.
Which tasks are most commonly AI-assisted?
Task-level use and satisfaction.
How much will developers let an agent do?
Autonomy today, and what blocks it.
Which developers should we target?
Use and awareness by role and team.
Numbers you can defend
- Unbranded. Respondents never know who commissioned the survey.
- Clean shares. Boosted samples stay out of share figures.
- Public. Anyone can check it. Read the methodology
About SlashData
SlashData is an analyst firm that researches developers and how AI is changing software development. We run large global developer surveys twice a year. For over 20 years, technology companies have used our research to make product and marketing decisions.
Trusted by the top technology companies
Be the first to see the data
Join the waitlist and we will email you when the results are out in October 2026.