# Claude Opus 3331 Hacker News commenters recommended Claude Opus and 2746 criticized it, November 2025 to November 2026. It was recommended in 13 of 13 months; owners who said how long report a median of 0.3 years (156 owners). **Verdict:** Recommended as a top coding model that fixes bugs and ships features with fewer mistakes than earlier LLMs, though critics cite weak architecture choices, usage limits and confident errors. ## Why people recommend it - Strong at coding and bug fixing (8 comments): "It's really good at fixing bugs in an established architecture or adding a similar feature, but it's absolutely mid tier at putting together a simple effective architecture" ([source](https://news.ycombinator.com/item?id=49217550)) "When Opus 4.5 entered the picture, there was a noticeable improvement in the way the LLM (for me), interacted with the code base, and the issues that it was able to find." ([source](https://news.ycombinator.com/item?id=47027758)) "Claude Code and Opus really took off when they improved the harnessing enough that it would self-correct many of its mistakes without needing user input" ([source](https://news.ycombinator.com/item?id=48627704)) - Fewer mistakes than other models (3 comments): "I don't recall i last few weeks where Opus 5.5 in CC made an obvious lie, made an unreasonable assumption, or made a mistake that smart human couldn't make." ([source](https://news.ycombinator.com/item?id=50048772)) "I've been using Claude Opus exclusively for while and it doesn't tend to make those mistakes nearly as often as I used to see with lower-powered LLMs." ([source](https://news.ycombinator.com/item?id=47838188)) "The frustrations with previous models going off the rails, getting stuck in loops, hacky solutions or bad code are almost completely gone with Opus." ([source](https://news.ycombinator.com/item?id=46562367)) - Big step up in capability (5 comments): "When Opus 4.5 entered the picture, there was a noticeable improvement in the way the LLM (for me), interacted with the code base, and the issues that it was able to find." ([source](https://news.ycombinator.com/item?id=47027758)) "I think Opus 4.5 & 4.6 are an impressive step up in capabilities but I'm really skeptical that this model is replacing the output of 9/10 skilled front end engineers" ([source](https://news.ycombinator.com/item?id=46972593)) "The frustrations with previous models going off the rails, getting stuck in loops, hacky solutions or bad code are almost completely gone with Opus." ([source](https://news.ycombinator.com/item?id=46562367)) - Good for research and analysis (3 comments): "For complex non-coding work (research, in-depth analysis, assembly of complex info-dense documents) I'd rather do it by hand than switch from Opus 4.7 to anything I could even theoretically run locally." ([source](https://news.ycombinator.com/item?id=48385004)) "if you tell a frontier LLM, like Opus 4.6, "Turn this paper into an interactive web page highlighting the most important aspects" it does a really good job" ([source](https://news.ycombinator.com/item?id=47197154)) "Claude web finds a number of small discrepancies in the sources, which I manually crosschecked and seem consistent with a human mixng things up slightly." ([source](https://news.ycombinator.com/item?id=48151253)) ## What people warn about - Confidently makes things up (5 comments): "It both confidently makes shit up, claims it is mathematically justified and when the resulting simulation explodes says "I apologise; I lied above"" ([source](https://news.ycombinator.com/item?id=48474666)) "it quickly fills much of the claude opus context window with relevant documentation alone, and even then repeatedly outputs things directly counter to the documentation it just ingested" ([source](https://news.ycombinator.com/item?id=46899348)) "I've recently seen Opus, after struggling for a bit, implement an API by having it return JSON that includes instructions for a human to manually accomplish the task I gave it." ([source](https://news.ycombinator.com/item?id=46585455)) - Poor architectural decisions (2 comments): "It's really good at fixing bugs in an established architecture or adding a similar feature, but it's absolutely mid tier at putting together a simple effective architecture" ([source](https://news.ycombinator.com/item?id=49217550)) "Opus and the like is amazing at Python, honestly sometimes better than me but it does do some really stupid architectural decisions occasionally." ([source](https://news.ycombinator.com/item?id=47373315)) - Usage limits and token costs (3 comments): "Even though I find Opus-es to be more well-rounded (and more useful) for certain tasks, I instinctively reach for ChatGPT / codex to avoid burning up my usage limits" ([source](https://news.ycombinator.com/item?id=46910883)) "Careless use, I guess. I was fixing a bunch of mocks in some once-great but now poorly maintained code, and I wasn't really feeling it so I just fed everything to Claude. Opus, unfortunately. I could easily have downgraded a bit." ([source](https://news.ycombinator.com/item?id=46543487)) "They reduced the cache TTL to one hour so if you leave your prompt sitting idle for an hour at 700,000 tokens the next time you hit enter send it it will be completely uncached and eat a ton of tokens." ([source](https://news.ycombinator.com/item?id=47810608)) - Guardrails block security work (3 comments): "Opus doesn't appear to refuse on simple keywords, but it does seem like Fable's reasoning introduces enough nefarious-sounding context that Opus will then refuse" ([source](https://news.ycombinator.com/item?id=49329995)) "Opus5.5 and other models are very limited by their own safety, I have digital twins of our infra and I want LLMs to actually write full exploits on it" ([source](https://news.ycombinator.com/item?id=50324477)) "I have been trying to work on debugging tools using Opus 4.8. As it turns out working with low-level techniques that inspects and alters behavior of other processes is bordering close enough to cybersecurity research that it often hits the guardrails." ([source](https://news.ycombinator.com/item?id=48527906)) ## Alternatives hackers recommend - [OpenAI GPT](https://hackerapproved.com/p/gpt-5-5.md): 1205 recommend - [Claude Sonnet](https://hackerapproved.com/p/claude-sonnet.md): 854 recommend - [DeepSeek V4 Flash](https://hackerapproved.com/p/deepseek-v4-flash.md): 828 recommend - [Fable](https://hackerapproved.com/p/fable.md): 823 recommend - [Claude Fable 5](https://hackerapproved.com/p/fable-5.md): 774 recommend ## Where to get it - [Visit the official site](https://claude.ai/login) ## Cite as Hacker Approved, "Claude Opus", https://hackerapproved.com/p/claude-opus/ (counts from 1,795,640 Hacker News comments, November 2025 to November 2026).