Skip to content
Open navigation
OpenOpen workspace

Stack Overflow Developer Survey 2026

AI coding agents: Used in the past year by work setting

In the Remote group, 76% selected “Claude Code”; in the In-person group, 64% did. Of the 33 answers tested, “Claude Code” differs most by work setting, a negligible association: χ²(4, N = 10323) = 99.21, Holm-adjusted p < .001, V = .10. Respondents could select more than one answer, so the percentages add to more than 100%. These are unweighted respondents, not population estimates.

Published by crosstabs.com · Last updated ·
How results are checkedNamed statistical review pending. Selected automated reference cases are published on the validation page.

The table

The table

Question asked: Have you used any of the following coding agents or assistants in the past year? Do you want to use any of the following in the next 6 months? Respondents could select more than one answer. Shown by work setting, as the percentage of each group's respondents who selected each answer, with the number of respondents beneath. A respondent is counted in every row they selected, so a column adds to more than 100%.

AI coding agents: Used in the past year by work setting: percentage of each group's respondents who selected each answer, and respondent counts
AI coding agents: Used in the past yearHybrid (some in-person, leans heavy to flexibility) (A)Hybrid (some remote, leans heavy to in-person) (B)In-person (C)Remote (D)Your choice (very flexible, you can come in when you want or just as needed) (E)All
Claude Code66.8%1,11969.1%C1,41663.7%89275.8%ABCE2,91068.3%93070.4%7,267
GitHub Copilot67.1%CDE1,12463.4%CD1,29957.2%80259.3%2,27761.6%83961.4%6,341
OpenAI Codex27.7%46329.2%59731.5%44133.7%ABE1,29229.2%39830.9%3,191
Cursor23.5%39324.7%50522.2%31132.0%ABCE1,22724.7%33626.9%2,772
Google Antigravity14.4%24115.5%31817.6%24616.2%62115.1%20515.8%1,631
Gemini Code Assist13.7%22913.0%26616.1%b22613.6%52314.2%19313.9%1,437
Open Code13.7%23012.7%26012.1%17014.4%55314.7%20013.7%1,413
JetBrains AI14.2%C23812.5%25710.6%14913.6%C52117.4%BCD23713.6%1,402
Windsurf6.3%1065.7%1166.4%897.7%B2975.8%796.7%687
Warp5.0%844.4%904.3%606.2%Bc2395.7%775.3%550
Cline4.3%724.1%835.6%795.0%1904.7%644.7%488
Pi4.0%673.9%793.1%435.1%C1954.8%664.4%450
Amazon Q Developer4.6%773.6%733.6%513.7%1415.4%bd744.0%416
Continue3.3%563.5%724.6%653.6%1404.6%633.8%396
Mistral Vibe4.1%D683.6%732.4%342.5%973.9%d533.1%325
Kilo Code3.2%542.4%492.9%413.4%1303.5%483.1%322
Devin2.9%482.5%511.9%273.4%c1312.9%392.9%296
Google Jules1.8%302.0%402.5%352.2%843.4%ab462.3%235
Aider2.0%341.5%301.4%202.6%b982.8%b382.1%220
Replit Agent1.9%321.5%301.9%262.3%902.6%352.1%213
Tabnine1.9%311.9%392.2%311.9%732.3%322.0%206
Codium (Qodo)1.3%211.5%301.3%181.6%622.5%341.6%165
BLACKBOX AI1.0%171.0%201.9%261.1%411.7%231.2%127
Supermaven1.1%191.0%201.3%181.2%471.5%211.2%125
Augment Code1.1%180.7%151.0%141.3%491.5%201.1%116
Respondents1,6742,0481,4013,8381,36210,323

A letter beside a percentage names the columns it is significantly higher than (uppercase 95%, lowercase 90%; pairwise column tests, Bonferroni-corrected). 20,580 of 30,903 respondents have no answer to one of the two questions (not asked or not answered) and are left out. The table shows 25 of the 33 answers: the most selected. The CSV and JSON downloads have every answer.

Analysis

Results

Chi-square test of “Claude Code”

χ² = 99.21

df = 4, Holm-adjusted p < .001

Effect size

Cramér's V = 0.098

a negligible association

The strongest answer, APA-style: χ²(4, N = 10323) = 99.21, Holm-adjusted p < .001, V = .10. Each answer is tested on its own, as the respondents who selected it against those who did not, across the 5 groups (N = 10,323). The p-values are Holm-adjusted for the 33 answers tested; 9 of them differ by work setting at the 0.05 level.

AI coding agents: Used in the past year by work setting: the test of each answer, selected against not selected, by work setting
AnswerSelectedχ² (df = 4)Adjusted pCramér's V
Claude Code70.4%99.21< .0010.098
GitHub Copilot61.4%44.05< .0010.065
OpenAI Codex30.9%26.91< .0010.051
Cursor26.9%84.65< .0010.091
Google Antigravity15.8%6.841.0000.026
Gemini Code Assist13.9%7.631.0000.027
Open Code13.7%7.411.0000.027
JetBrains AI13.6%29.73< .0010.054
Windsurf6.7%12.58.2170.035
Warp5.3%13.33.1760.036
Cline4.7%5.761.0000.024
Pi4.4%12.91.2000.035
Amazon Q Developer4.0%11.30.3510.033
Continue3.8%6.791.0000.026
Mistral Vibe3.1%15.46.0840.039
Kilo Code3.1%5.461.0000.023
Devin2.9%9.60.6210.030
Google Jules2.3%10.59.4420.032
Aider2.1%13.88.1540.037
Replit Agent2.1%7.351.0000.027
Tabnine2.0%1.651.0000.013
Codium (Qodo)1.6%9.36.6210.030
BLACKBOX AI1.2%9.42.6210.030
Supermaven1.2%2.341.0000.015
Augment Code1.1%5.321.0000.023

How to read it

No single chi-square test covers this table. A respondent appears in every row they selected, so the rows are not separate groups of people, and a test of the whole table would count the same person several times. Each answer is tested on its own instead: did the share who selected it differ between the groups? Testing 33 answers gives 33 chances of a false alarm, so the p-values are Holm-adjusted to keep the chance of any false alarm in this table at 0.05.

For “Claude Code” the test rejects independence at the 0.05 level (Holm-adjusted p < .001): differences this large between the groups are unlikely if selecting it did not depend on work setting. Cramér's V of 0.098 is in the negligible range, which is trivially weak in practical terms, even where a sample this large can detect it.

With 10,323 respondents, even small differences reach statistical significance. Read the percentages and the effect size before the p-value.

Limits

  • The survey has no weights. These tables are unweighted counts of the people who answered, so they describe the respondents, not developers in general.
  • The test treats respondents as a simple random sample. They are people who chose to answer a survey Stack Overflow promoted, not a random sample of developers, so a result holds for these respondents and may not hold beyond them.
  • The adjustment covers the answers of this question by this group only, not the other tables in the library. The headline answer is the strongest of those tested, so its effect size is larger than a typical answer's.
  • A difference between groups is not evidence that group membership causes the answer.

Next steps

Work with this table

Open the test of “Claude Code” in the calculator

The link fills in the respondents who selected this answer and those who did not, by group. The whole table does not open there: its rows overlap, and the calculator's test needs rows that are separate people.

Open this question in the full file

The workspace opens all 30,903 respondents in your browser with this question selected. The survey has no weights, so the table there is unweighted too. Change the group, add a filter or export from there.

Open in the workspace →

Run this on your own data — free, no signup

Upload a CSV, Excel, SPSS, Stata, SAS, R, Parquet or JSON file. Everything runs in your browser; your file never leaves your device.

Open the workspace →

The same question by other groups

Other questions that differ by work setting

Source

Stack Overflow. 2026 Developer Survey, https://survey.stackoverflow.co/2026, licensed under the ODbL 1.0. Tabulated by crosstabs.com. Variables AICodeAgentHaveWorkedWith (AI coding agents: Used in the past year) and WorkSetting. Stack Overflow licenses the survey data under the Open Database License (ODbL) 1.0 and its contents under the Database Contents License (DbCL) 1.0: credit Stack Overflow, keep the licence notice, and share any database made from the data under the ODbL. These tables and their downloads are offered on those terms.

  1. [1]Stack Overflow: 2026 Developer Survey public results (survey_results_public.csv)survey.stackoverflow.co
  2. [2]Stack Overflow: Open Database License (ODbL) 1.0opendatacommons.org