OpenAI Codex vs JetBrains Junie: what developers actually say
Based on 2,066 opinions on Hacker News from developers who discussed these two, most of them first-hand users.
OpenAI Codex comes out ahead. It has far more evidence at 2,013 opinions from 1,386 authors, and it leads on code quality (343 positive to 116 negative), cost (174 to 54), and autonomy (169 to 46). Its reliability mentions are net negative (112 positive, 127 negative), usage limits are mixed (131 to 61), UX is positive but softer (131 to 86), and 90-day approval is 63%. JetBrains Junie is the better pick for developers already in JetBrains IDEs who want cheap help on small refactors; its 53 opinions are too few for a confident overall call, and its 90-day approval is 33%.
When developers explicitly said they prefer one or switched: 1 picked OpenAI Codex, 0 picked JetBrains Junie.
Choose OpenAI Codex if
- Want a terminal agent with a large evidence base
- Care about code quality, cost, and autonomy
- Can accept mixed reliability and quota complaints
Choose JetBrains Junie if
- Work inside JetBrains IDEs all day
- Want low-cost help on small refactors
- Mostly need simple tasks done predictably
Where each one wins and loses
Why developers switched
Picked OpenAI Codex
Picked JetBrains Junie
No explicit switches recorded.
Summaries are paraphrased. Each links to the original comment.