# Claude Fable 5 774 Hacker News commenters recommended Claude Fable 5 and 1004 criticized it, November 2025 to November 2026. It was recommended in 6 of 13 months; owners who said how long report a median of 0.01 years (22 owners). **Verdict:** Recommended for long-horizon agentic work and hard reasoning, debugging and research problems where it beats Opus, though overbroad safety filters often block legitimate security and biology tasks. ## Why people recommend it - Excels at long-horizon agentic tasks (6 comments): "Useful interactively in coding agent but boy if you have a kind of task that Fable can do in an agentic loop that Opus can't then you're in a good place." ([source](https://news.ycombinator.com/item?id=48742453)) "Fable was in a different league. Especially when tasked with long horizon goals that involved reasoning at a high and low level to solve the task." ([source](https://news.ycombinator.com/item?id=48513107)) "I had good experiences doing multi-hour refactoring/housekeeping tasks that basically consisted of applying the same steps and rules n times." ([source](https://news.ycombinator.com/item?id=48498575)) - Cracks hard, deep technical problems (8 comments): "Fable 5.1 found the cause of a rare crash on their internal systems that none of their engineers (or any other model) had been able to explain" ([source](https://news.ycombinator.com/item?id=49525815)) "this feels like the difference between having a post-doctoral colleague and educating a student that I have to constantly guide and correct" ([source](https://news.ycombinator.com/item?id=48535352)) "it was pretty impressive that it could consume four pages of KiCad schematic and reconstruct the full topology and theory of operation" ([source](https://news.ycombinator.com/item?id=48498881)) - Clear upgrade over Opus (4 comments): "This has been a much better rollout. The tool calling is not broken out of the gate like 4.8 was, and the tokens generation is fast." ([source](https://news.ycombinator.com/item?id=48470920)) "it might be much better on all dimensions (speed, cost, quality) to just use Fable 5 on low/medium effort than switch to Opus" ([source](https://news.ycombinator.com/item?id=48464394)) "Fable on "high" is producing substantially better results than Open 4.8 on xhigh for me and my actual real-world evals today." ([source](https://news.ycombinator.com/item?id=48468707)) - Pushes back instead of flattering (2 comments): "you should probably be doing this discussion work along with Fable 5. It will give good feedback if you're working on the correct things." ([source](https://news.ycombinator.com/item?id=49217296)) "I don't get any sycophancy from it. The conversations are actually quite good, and it'll push back if it thinks I'm wrong." ([source](https://news.ycombinator.com/item?id=48974315)) ## What people warn about - Overbroad safety filters flag legitimate work (8 comments): "these guardrails are counter productive for well intentioned cyber security, because I can't use it to test and harden my own software" ([source](https://news.ycombinator.com/item?id=48492846)) "Claude's response has been degraded for innocuous research projects concerning population-level analyses of astronaut health." ([source](https://news.ycombinator.com/item?id=49526114)) "Fable 5's safeguards flagged this message. Our intentionally broad safeguards deliver more capabilities but can also flag safe coding, cybersecurity, and biology tasks." ([source](https://news.ycombinator.com/item?id=49003710)) - Silent degradation and hidden prompt changes (4 comments): "Fable post-ban was never the same. At the least, whatever system prompt munging or pre/post filtering they did to strengthen the guardrails nerfed it." ([source](https://news.ycombinator.com/item?id=49790752)) "this is backfiring in many ways including safety of your code because it may fail successfully if your context contains something they don't like" ([source](https://news.ycombinator.com/item?id=48499983)) "Unlike our interventions for cybersecurity, biology and chemistry, and distillation attempts, these safeguards will not be visible to the user." ([source](https://news.ycombinator.com/item?id=48466251)) - Tight usage limits, heavy token burn (5 comments): "Limiting it to half the weekly limit stings, but if you have a quota reset in the middle of it you still get effectively one weekly limit." ([source](https://news.ycombinator.com/item?id=48753029)) "The writing style has significantly improved, however the token burn rate for tasks I have been working on seems to have skyrocketed." ([source](https://news.ycombinator.com/item?id=49530347)) "Still does not crack my hardest nuts. Gave it one of them and it blew through my entire allowance on thinking about one question" ([source](https://news.ycombinator.com/item?id=48468599)) - Not better for everyday coding (5 comments): "I think the results wouldn't be so different if you manually challenged 4.6 with enough iterations of logs, screenshots, and follow up questions." ([source](https://news.ycombinator.com/item?id=48501134)) "Fable actually returned a result that fails and confidently stated it ran X, Y, Z tests to ensure it works and got these results." ([source](https://news.ycombinator.com/item?id=48495500)) "Fable seems to be a larger model. It costs more to run and does not seem superior for _typical_ software engineering work." ([source](https://news.ycombinator.com/item?id=48960177)) ## Alternatives hackers recommend - [Claude Opus](https://hackerapproved.com/p/claude-opus.md): 3331 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 ## Cite as Hacker Approved, "Claude Fable 5", https://hackerapproved.com/p/fable-5/ (counts from 1,795,640 Hacker News comments, November 2025 to November 2026).