Claude Opus 4.5 vs Claude Opus 4.6: what 11 developers chose.
609 opinions. 11 developers named a direct preference: Claude Opus 4.5 7–4. Approval is 78% vs 63%.
Claude Opus 4.5 wins the direct preference 7–4, but both tools score best on the same topic. Use the decision cards below to choose on workflow.
Choose Claude Opus 4.5 if you…
- Want higher approval (78 vs 63)
- Care about code quality sentiment (120 positive vs 28 negative)
- Prefer short, instruction-heavy interactive sessions
Choose Claude Opus 4.6 if you…
- Need better following of negative instructions
- Weight UX comments, where it leads 4 to 1
- Want the Opus release 12 users ranked above Opus 4.7
Claude Opus 4.5 and Claude Opus 4.6: where each one wins.
Net sentiment per topic: praise minus complaints, as a share of mentions. Positive means more praise than complaints. Hover a dot for the raw counts.
Net sentiment = praise minus complaints, as a share of each tool's mentions. The dot lanes are only for readability; position is the score.
View data table
| Topic | Claude Opus 4.5 | Claude Opus 4.6 | Mentions |
|---|---|---|---|
| Code quality | +62 | +29 | 148 / 121 |
| Reliability | +3 | -34 | 35 / 58 |
| Autonomy | +76 | +45 | 34 / 29 |
| Large codebases | +79 | 0 | 19 / 8 |
| Cost & pricing | +67 | 0 | 12 / 14 |
| Speed | +100 | -20 | 15 / 10 |
| UX | +43 | +60 | 7 / 5 |
| Usage limits | +20 | +14 | 5 / 7 |
The same decision, from both sides.
Paraphrased from comments where a developer named a preference or described a switch. Each links to the original.
Picked Claude Opus 4.5 7 developers
Said 4.6 overthought tasks and spun for long stretches while 4.5 stayed responsive, so they skipped later releases.
Argued Anthropic tuned 4.6 for cost rather than capability and called it a bad model after 4.5 handled longer tasks well.
Preferred 4.5 for fine-grained instruction workflows and saw 4.6 as a regression tuned for long-running, less interactive tasks.
Found 4.6 post-trained to patch code quickly, producing sloppy fixes that showed it did not understand the problem.
Picked Claude Opus 4.6 4 developers
Said 4.6 was a significant step up from 4.5 and that staying on 4.5 would have meant worse results and slower shipping.
>> November 2024 through February 2026 Yeah, listen... I'm glad these types of studies are being conducted. I'll say this though: the difference between pre- and post-Opus 4.5 has been night and day for me. From August 2025 through November 2025 I led a complex project at work where I used Sonnet 4.5 heavily. It was very helpful, but my total productivity gains were around 10-15%, which is pretty much what the study found. Once Opus came out in November though, it was like someone flipped a switch. It was much more capable at autonomous work and required way less hand-holding, intervention or
Said 4.5 improved sharply over earlier models and 4.6 improved again, with the pace of gains increasing.
Found 4.6 better than 4.5 at following instructions, especially negative rules, and at applying skills.
The numbers.
| Metric | Claude Opus 4.5 | Claude Opus 4.6 |
|---|---|---|
| Approval | 78% | 63% |
| Last 90 days | 90% | 63% |
| Opinions | 322 | 287 |
| Developers | 259 | 257 |
| First-hand opinions | 254 | 242 |
| Direct picks | 7 | 4 |
| Type | Model | Model |
| Made by | — | — |
Related matchups.
Claude Opus 4.6 vs GPT-5.4
10–4 direct picks, and a clear difference in what developers optimise for.
Claude Opus 4.6 vs Claude Opus 4.7
12–2 direct picks, and a clear difference in what developers optimise for.
Claude Opus 4.5 vs GPT-5.2
7–3 direct picks, and a clear difference in what developers optimise for.
How to read this comparison.
Approval is the positive share of opinions that took a side, with the rank based on the lower bound of a 90% confidence interval. Direct picks count comments where a developer explicitly preferred one tool or described switching. Small samples are noisier; check the opinion counts.
Get the monthly report
Which AI coding tools gained or lost developer approval on Hacker News, and why people switched. One email a month.