Skip to content
unnudged
Get early access

Catalog · Category

AI coding agents and editors

13 topics · 4,853 posts in the last 7 days

Cursor

brand

People still credit Cursor with real shipped work and recommend it by name, but the argument has moved to what the bundled X plan gives them and what happens when something breaks.

Rising
24h
280
7 days
1,392
Momentum
+41%

Claude Code

brand

The biggest single theme is finished work: people describe apps, automations and fixes Claude Code completed, often naming the setup that made it work.

Rising
24h
176
7 days
855
Momentum
+44%

OpenAI Codex

brand

The biggest share of the talk is people showing what Codex builds for them.

Cooling
24h
69
7 days
773
Momentum
−38%
  1. 1CursorAI coding agents and editors280+41%
  2. 2Claude CodeAI coding agents and editors176+44%
  3. 3OpenAI CodexAI coding agents and editors69−38%
  4. 4Cognition (Devin)AI coding agents and editors4−94%
  5. 5GitHub CopilotAI coding agents and editors70+90%
  6. 6Gemini CLIAI coding agents and editors——
  7. 7JetBrainsAI coding agents and editors60+116%
  8. 8AmpAI coding agents and editors43+61%
  9. 9ZedAI coding agents and editors10−57%
  10. 10FactoryAI coding agents and editors4−82%
  11. 11KiroAI coding agents and editors0−100%
  12. 12WarpAI coding agents and editors7−36%
  13. 13Kilo CodeAI coding agents and editors1−50%

How they compare

OpenAI Codex has the largest conversation here, with 568 people on its page. For each brand, this is the largest group of people behind each question, counted when its page was written. Open a brand to read the posts behind the numbers.

BrandTop reactionTop complaintTop wishMost compared with
OpenAI Codex568 peopleTop reactionEnthusiastic about what it builds 112 peopleTop complaintSloppy or heavy-handed output 28 peopleTop wishAgents that hand off to each other 9 peopleMost compared withClaude Code 74 people
Cursor544 peopleTop reactionPraises it 107 peopleTop complaintBundled plan feels like a downgrade 33 peopleTop wishOne shared rules file 5 peopleMost compared withClaude Code 30 people
Claude Code359 peopleTop reactionWorks well 150 peopleTop complaintExpensive 23 peopleTop wishShared memory across agents 3 peopleMost compared withCodex 42 people
Amp239 peopleTop reactionLoves it 93 peopleTop complaintBugs and rough edges 19 peopleTop wishMultiple subscriptions at once 4 peopleMost compared withT3 Code 5 people
Zed151 peopleTop reactionLoves Zed 58 peopleTop complaintBugs and broken features 5 peopleTop wishLaTeX and math rendering 4 peopleMost compared withVS Code 11 people
JetBrains149 peopleTop reactionLoves the IDEs 29 peopleTop complaintSubscription cost and renewals 6 peopleTop wish—Most compared withVS Code 9 people
GitHub Copilot101 peopleTop reactionPraises it 35 peopleTop complaintPrice too high 9 peopleTop wishProduce docs and harnesses too 3 peopleMost compared withCursor 4 people
Cognition (Devin)95 peopleTop reactionLoves it 27 peopleTop complaintCrashes and lost sessions 5 peopleTop wishShow the numbers behind the claims 5 peopleMost compared withClaude Code 9 people
Factory86 peopleTop reactionFans, plain and loud 47 peopleTop complaintUsage runs out too fast 8 peopleTop wishMore generous limits 4 peopleMost compared withClaude Code 3 people
Gemini CLI72 peopleTop reactionPositive but measured 15 peopleTop complaintRenames and replacements 9 peopleTop wishClear naming and one place to look 2 peopleMost compared withAntigravity 2 people
Kiro61 peopleTop reactionPositive word of mouth 11 peopleTop complaintModel credits cost too much 4 peopleTop wishAdd Haiku 5.5 as a model 4 peopleMost compared withCursor and Copilot 2 people
Warp49 peopleTop reactionLoves it 13 peopleTop complaintBugs and regressions 4 peopleTop wishAgents should prove work with tests and screenshots 2 peopleMost compared withGhostty 2 people
Kilo Code18 peopleTop reactionEnthusiastic 4 peopleTop complaintFreezes, login and server errors 3 peopleTop wish—Most compared with—