AI coding agents · scenario rankings

Inside an IDE / editor

Most Popular AI Coding Agents Inside an IDE

Concentrated

Measured 2026-09-22 · 12-seat model panel · 10 scenarios · methodology #2 v2026-09-22.1

Download image (with QR code)Share on X Share on LinkedIn

WeChat: save the image and long-press the QR code in a chat.

The ten

Rank	Entity	Mention rate (of answers)	Wilson 95% interval	Seats that mentioned it	Seats with stable coverage
1	GitHub Copilot	99.8%	[98.6%, 100.0%]	12 of 12	12 of 12
2	Windsurf (Devin Desktop)	99.5%	[98.3%, 99.9%]	12 of 12	12 of 12
3	Cursor	95.7%	[93.2%, 97.2%]	12 of 12	12 of 12
4	Tabnine	84.3%	[80.5%, 87.5%]	12 of 12	12 of 12
5	Amazon Q Developer	81.0%	[76.9%, 84.5%]	12 of 12	11 of 12
6	Continue.dev	76.4%	[72.1%, 80.2%]	12 of 12	10 of 12
7	JetBrains AI	70.8%	[66.3%, 75.0%]	12 of 12	10 of 12
8	Cline	69.6%	[65.1%, 73.9%]	12 of 12	9 of 12
9	Sourcegraph Cody	48.0%	[43.2%, 52.8%]	12 of 12	6 of 12
10	Replit	38.1%	[33.5%, 42.8%]	11 of 12	4 of 12

How it is counted: mention rate = answers naming it ÷ 415 (7 phrasings × 5 repeats × 12 seats = 420 calls, 415 valid); the last two columns count seats, and "stable coverage" means more than half of that seat's 35 answers. Different denominators, so 12 of 12 next to a rate below 100% is not a contradiction.

The sample behind this cell

Valid samples / calls expected: 415 / 420
Completeness: 98.8%
Effective entities (effN): 14.5
Entities mentioned at least once: 67

How this cell was asked

This cell used 7 phrasings, each asked 5 times. Every call asked for the ten names only — no reasons, no scores.

Most Popular AI Coding Agents Inside an IDE
What are the best AI coding agents that live inside an editor or IDE?
Which in-editor agentic coding assistants are people using these days?
I want an AI agent that works inside my code editor and can make multi-file changes. What should I look at?
Looking for an AI-native IDE or editor extension with an autonomous coding agent. What are the options?
best ai coding agent for ide editor extension
List some AI coding agents that run inside an IDE or editor.

Where this page stops

This is one measurement, not a trend line. The next run of the same phrasings comes next quarter.
Each call collected ten names at most. Anything past the tenth name is invisible in this cell, which is not the same as being absent.
Mention rate is the share of samples in this cell where a name came up. It is not market share, and it is not a judgement about the product.

Methodology: the panel, the phrasings and how completeness is counted

You can also query this data from Claude, Cursor or another agent, with the same reading rules attached. For agents →