OpenAI Codex: loved for code quality, held back by reliability.
2,013 Hacker News opinions from 1,386 developers, Nov 2025 – Nov 2026. 1,708 come from people who say they use it. Here is what they praise, what they complain about, and who switched.
Where OpenAI Codex is strong — and where it is not.
OpenAI Codex earns 69% approval across 2,013 opinions. Developers praise its code quality, while reliability is the recurring complaint. In the last 90 days it sits at 63%.
Pick it if you
- Want strong code quality and hands-off autonomy
- Like starting agent sessions and doing other work meanwhile
- Value Codex billing kept separate from chat on a Pro plan
Look elsewhere if you
- Need dependable execution on every single run
- Expect generous weekly quotas on the $20 plan
- Depend on a polished TUI or fast Windows performance
Where it wins and loses.
Mentions across 2,013 opinions. Bars are counts, not votes; a tool can be both praised and criticized on the same topic.
View data table
| Topic | Praise | Complaints | Mentions |
|---|---|---|---|
| Code quality | 343 | 116 | 459 |
| Autonomy | 169 | 46 | 215 |
| Cost & pricing | 174 | 54 | 228 |
| Usage limits | 131 | 61 | 192 |
| Speed | 96 | 33 | 129 |
| UX | 131 | 86 | 217 |
| Customization | 33 | 18 | 51 |
| Reliability | 112 | 127 | 239 |
How opinion has moved.
Monthly approval, Nov 2025 – Oct 2026. Bars show opinion volume; low-volume months are still included but faded.
View data table
| Month | Approval | Opinions |
|---|---|---|
| Nov 2025 | 61% | 101 |
| Dec 2025 | 65% | 143 |
| Jan 2026 | 70% | 147 |
| Feb 2026 | 70% | 174 |
| Mar 2026 | 78% | 181 |
| Apr 2026 | 76% | 202 |
| May 2026 | 71% | 207 |
| Jun 2026 | 76% | 190 |
| Jul 2026 | 65% | 162 |
| Aug 2026 | 73% | 196 |
| Sep 2026 | 55% | 173 |
| Oct 2026 | 59% | 128 |
What developers say.
Summaries are paraphrased. Each links to the original Hacker News comment.
Praise
4 selectedReports agents on a Codex Max subscription built a reasonable RCT2 clone in about 48 hours.
They start Codex sessions and do other things while it works.
OpenAI's Pro plan bills Codex separately from chat, which made it valuable for research work.
Prefers Codex's desktop app over the TUI and says the Linux desktop app works fine.
Complaints
4 selectedCodex Astra 6 misses details and needs rewrites, so humans are more trustworthy on some tasks.
Full-time user who dislikes the TUI experience of these agents and is building a GUI instead.
Native subagents in Codex hit usage limits, so delegating work to other harnesses helps.
Burned through the weekly quota quickly; finds OpenAI's allowance much less generous than Claude's.
Questions developers ask.
Is OpenAI Codex worth it?
Is Codex better than Claude Code?
How are Codex usage limits and pricing?
The data behind this page.
Developers give OpenAI Codex a 69% approval score across 2,013 opinions from 1,386 authors, with 1,708 opinions from firsthand users. Code quality is the strongest aspect at 343 positive mentions against 116 negative, and autonomy (169 to 46) and cost (174 to 54) also score well. Reliability is the weak spot at 112 positive against 127 negative. Approval peaked at 78 in March 2026, fell to 55 in September 2026, and stood at 59 in October; the 50 for November rests on only 9 opinions. In direct comparisons, 181 developers prefer Codex over Claude Code while 62 prefer Claude Code instead.
Every number comes from public Hacker News comments collected through the HN Search API. A language model tagged each comment for sentiment, first-hand use, topic and switches; a human spot-checks samples. No vendor can pay for placement.
Get the monthly report
Which AI coding tools gained or lost developer approval on Hacker News, and why people switched. One email a month.