{"slug": "claude-fable", "name": "Claude Fable", "category": "ai", "category_name": "AI models and assistants", "description": "A Claude AI model from Anthropic used for coding, planning and orchestrating agents.", "recommenders": 175, "commenters": 372, "warners": 91, "score": 100.256, "rising": 46.95, "first": "2026-03-20", "latest": "2026-10-29", "why": "Commenters like Fable as a planner and orchestrator that directs cheaper subagents such as Opus, and for reliable code on hard problems.", "why_is_quote": false, "has_verdict": true, "sentiment": {"recommend": 182, "neutral": 89, "compare": 36, "warn": 100}, "spark": [0, 0, 0, 0, 1, 0, 0, 44, 56, 26, 44, 11], "listed": true, "loved": ["Planning and orchestration", "Code quality"], "caveat": "Cost and limits", "short_quote": "it is really good, definitely a step change from Opus 4.8, both for coding and general chat-style discussions.", "icon": true, "domain": "fable.co", "github_repo": null, "official_url": "https://fable.co", "pricing": "paid", "aliases": ["Fable", "Fable 5", "Claude Fable", "Claude Fable 5", "Fable 5.1", "Claude Fable 5.1"], "verdict": [{"text": "Commenters like Fable as a planner and orchestrator that directs cheaper subagents such as Opus,", "cites": [{"id": 796868720, "item_id": 49804295, "url": "https://news.ycombinator.com/item?id=49804295", "by": "port3000", "date": "2026-09-22", "stance": "recommend", "quote": "I use Fable as the orchestrator, only speak with Fable", "excerpt": "I believe Opus 5 isn't meant to be spoken to by humans. It's great at executing but I reckon it's intended to be spoken to by other models such as Fable. I use Fable as the orchestrator, only speak with Fable, and all implementation, recon, design etc happens with Opus 5, with Fable reviewing (and translating).", "context": "", "use_case": "orchestrating coding agents", "switched_from": "", "story": "Claude Opus 5.5"}, {"id": 796896816, "item_id": 49806051, "url": "https://news.ycombinator.com/item?id=49806051", "by": "hombre_fatal", "date": "2026-09-22", "stance": "recommend", "quote": "My main process is to use Fable to reason and then spawn Opus subagents, and I get amazing results", "excerpt": "I mainly use Codex/Sol to review my plans drafted by Fable. But beyond that, Astra blows through usage limits too fast to be a daily driver and writes weird code despite what my \"house style\" is, and Codex is behind Claude Code in terms of critical features like seeing what's going on in subagents. The parent + subagent workflow has become critical for keeping the reasoning agent (parent) context-lean while also lett", "context": "", "use_case": "planning and reasoning", "switched_from": "", "story": "GPT-6 Sol and Luna"}, {"id": 794875328, "item_id": 49679708, "url": "https://news.ycombinator.com/item?id=49679708", "by": "pdantix", "date": "2026-09-13", "stance": "recommend", "quote": "having fable plan and orchestrate with opus implemention + fable reviews, is my go-to.", "excerpt": "personally, i would not rely on opus 5 end to end as it'll start getting into walls of comment slop and shitting up the codebase similar to gpt 5.5's isRecord meme. on the other hand, having fable plan and orchestrate with opus implemention + fable reviews, is my go-to. if you give fable your guidelines up front or in your {claude,agents}.md, it will keep opus on a tight leash. opus can still write great code almost ", "context": "", "use_case": "planning and reviewing code", "switched_from": "", "story": "Real-SWE: Benchmarking AI models on private, real-world, enterprise codebases"}, {"id": 794771361, "item_id": 49673210, "url": "https://news.ycombinator.com/item?id=49673210", "by": "mariocesar", "date": "2026-09-12", "stance": "recommend", "quote": "I now use Fable to create a PLAN.md.", "excerpt": "I have an 'ask' alias in my shell that just uses Haiku. I use it daily for pretty much everything For more long work, I now use Fable to create a PLAN.md. I tell it to make a plan that will be executed by other models, and most of the time it ends up choosing Opus or Sonnet. I didn't start doing this recently. Before that, I would just use the top model for everything. Splitting the work across different models depen", "context": "", "use_case": "planning long work", "switched_from": "", "story": "Ask HN: What default model do you use and why?"}]}, {"text": " and for reliable code on hard problems.", "cites": [{"id": 794870257, "item_id": 49679391, "url": "https://news.ycombinator.com/item?id=49679391", "by": "qskousen", "date": "2026-09-13", "stance": "recommend", "quote": "Fable 5.1 was noticeably better at getting things right and doing things in a good reliable way.", "excerpt": "I have a pet project I have been working away on for some time that involves building GPU backends for various cards in Zig, lots of complex stuff in it. Lately I mostly use Opus 5, it can pretty reliably plug away at things but it does mess stuff up occasionally. For this codebase, Fable 5.1 was noticeably better at getting things right and doing things in a good reliable way. Of course, I can only use Fable for a b", "context": "", "use_case": "tough coding tasks", "switched_from": "", "story": "Everyone should slow down AI development except for me"}, {"id": 801351969, "item_id": 50084498, "url": "https://news.ycombinator.com/item?id=50084498", "by": "jwr", "date": "2026-10-14", "stance": "recommend", "quote": "With newer models, and especially large ones like Fable, it just doesn't happen anymore.", "excerpt": "> Every time I stop reviewing the code from Claude I later find janky crap I'm very curious about that because that has not been my experience at all. I used to find janky crap from older models and I had to review every line of code very carefully. With newer models, and especially large ones like Fable, it just doesn't happen anymore. The code that I get is what I would have written. Actually, it's usually better b", "context": "", "use_case": "AI coding", "switched_from": "", "story": "Redesigning source control at Anthropic"}, {"id": 794786592, "item_id": 49674162, "url": "https://news.ycombinator.com/item?id=49674162", "by": "jamesponddotco", "date": "2026-09-12", "stance": "recommend", "quote": "Fable is the only one that fucking listens to me and writes proper Go code", "excerpt": "Given that Fable is the only one that fucking listens to me and writes proper Go code, that\u2019s what I use for everything code-related. At least on my personal projects. If I don\u2019t care about the code, i.e. I\u2019m writing something quick just to test something, I\u2019ll use GPT Astra to save my Fable tokens. For conversations and research I usually go with GPT Astra Pro. For Home Assistant I use a combination of Grok 4.20 and", "context": "", "use_case": "code", "switched_from": "", "story": "Ask HN: What default model do you use and why?"}]}, {"text": " The most common complaint is that it is expensive and uses up limits quickly,", "cites": [{"id": 796647731, "item_id": 49790483, "url": "https://news.ycombinator.com/item?id=49790483", "by": "swalsh", "date": "2026-09-21", "stance": "warn", "quote": "Fable chewed through my usage too quickly to be practical", "excerpt": "Codex has become my goto tooling. I used to be a Claude Max subscriber, but I was becoming disappointed with the quality of the output from Opus 5. Fable chewed through my usage too quickly to be practical. Moving to a Pro account w/ Codex was a big improvement. Sol had great output, and the usage was more than sufficient for most of my needs. However astra does tend to chew up usage, so when i've done to much of tha", "context": "", "use_case": "", "switched_from": "", "story": "Grok 4.7"}, {"id": 796957489, "item_id": 49809843, "url": "https://news.ycombinator.com/item?id=49809843", "by": "matheusmoreira", "date": "2026-09-22", "stance": "warn", "quote": "Anthropic has a separate meter for Fable. I used to get like five Fable sessions per week and that's it.", "excerpt": "Anthropic has a separate meter for Fable. I used to get like five Fable sessions per week and that's it. OpenAI has no such nonsense. No separate meter. No five hour limits. I get to use Astra at max effort on literally every task if I want to, and even this somehow lasts me several days. Anthropic got caught playing stupid \"20x refers to the 5h limit\" word games with their customers. Meanwhile, I have statistically ", "context": "former Anthropic customer", "use_case": "coding", "switched_from": "", "story": "GPT-6 Sol and Luna"}, {"id": 796955810, "item_id": 49809738, "url": "https://news.ycombinator.com/item?id=49809738", "by": "schipperai", "date": "2026-09-22", "stance": "recommend", "quote": "Fable is on a class of its own when it comes to coding and orchestration", "excerpt": "I have used Fable, Astra, Opus 5 and Sol 5.6 in Claude Code and Codex heavily since each of these have been released. Recently, I settled for Sol 5.6 as my daily driver, but with this release Opus 5.5 might be the one. Fable is on a class of its own when it comes to coding and orchestration, but it runs out pretty quick and is prohibitively expensive and slow. For me, Astra wasn't as good of an upgrade from Sol when ", "context": "", "use_case": "coding and orchestration", "switched_from": "", "story": "Claude Opus 5.5"}, {"id": 794842177, "item_id": 49677636, "url": "https://news.ycombinator.com/item?id=49677636", "by": "ricardobeat", "date": "2026-09-12", "stance": "warn", "quote": "I don\u2019t have a use for Fable at current prices.", "excerpt": "Opus 5 at low effort is quite good at most tasks, and will go a long way even on the base plan. At higher thinking levels it is unbearably cautious and verbose - it drives me mad. I don\u2019t have a use for Fable at current prices. My current go-tos are DS Flash, Minimax M3 for subagents. Adding Muse Spark 1.3 to that mix but not convinced yet.", "context": "", "use_case": "", "switched_from": "", "story": "Ask HN: What default model do you use and why?"}]}, {"text": " and some also dislike its writing style.", "cites": [{"id": 804494592, "item_id": 50280912, "url": "https://news.ycombinator.com/item?id=50280912", "by": "n0on3", "date": "2026-10-28", "stance": "warn", "quote": "Fable 5.1 choice of words and general \u201cwords salad\u201d expression are horrific.", "excerpt": "Speaking for a user-facing use, not autonomous coding or agentic workflows: Fable 5.1 choice of words and general \u201cwords salad\u201d expression are horrific. It starts not so good, then as the context grows (still way below any reasonable rotting threshold) it gets completely insane. It\u2019s pointless for a model to be extremely capable if it gives you brain damage. I don\u2019t know if it\u2019s because of the watermark thingy, all I", "context": "", "use_case": "user-facing writing", "switched_from": "", "story": "Gemini 4 Carbon"}, {"id": 796864993, "item_id": 49804062, "url": "https://news.ycombinator.com/item?id=49804062", "by": "visarga", "date": "2026-09-22", "stance": "warn", "quote": "The poor writing style made even models like Fable look worse than Sonnet 3.7 to me.", "excerpt": "I tried a lot to get rid of Claude vibes and I failed. I made blacklists, I used prompt evolution to test against good human writing samples, even created a dynamic policy for dialogue, a cli tool used to steer agents dynamically by well chosen questions. The poor writing style made even models like Fable look worse than Sonnet 3.7 to me.", "context": "", "use_case": "writing", "switched_from": "", "story": "I don't want to read what you didn't write"}]}], "praise": [{"label": "Planning and orchestration", "summary": "Many use it to plan and review while Opus, Sonnet or cheaper models do the implementation.", "quotes": [{"id": 796868720, "item_id": 49804295, "url": "https://news.ycombinator.com/item?id=49804295", "by": "port3000", "date": "2026-09-22", "stance": "recommend", "quote": "I use Fable as the orchestrator, only speak with Fable", "excerpt": "I believe Opus 5 isn't meant to be spoken to by humans. It's great at executing but I reckon it's intended to be spoken to by other models such as Fable. I use Fable as the orchestrator, only speak with Fable, and all implementation, recon, design etc happens with Opus 5, with Fable reviewing (and translating).", "context": "", "use_case": "orchestrating coding agents", "switched_from": "", "story": "Claude Opus 5.5"}, {"id": 794875328, "item_id": 49679708, "url": "https://news.ycombinator.com/item?id=49679708", "by": "pdantix", "date": "2026-09-13", "stance": "recommend", "quote": "having fable plan and orchestrate with opus implemention + fable reviews, is my go-to.", "excerpt": "personally, i would not rely on opus 5 end to end as it'll start getting into walls of comment slop and shitting up the codebase similar to gpt 5.5's isRecord meme. on the other hand, having fable plan and orchestrate with opus implemention + fable reviews, is my go-to. if you give fable your guidelines up front or in your {claude,agents}.md, it will keep opus on a tight leash. opus can still write great code almost ", "context": "", "use_case": "planning and reviewing code", "switched_from": "", "story": "Real-SWE: Benchmarking AI models on private, real-world, enterprise codebases"}, {"id": 794771361, "item_id": 49673210, "url": "https://news.ycombinator.com/item?id=49673210", "by": "mariocesar", "date": "2026-09-12", "stance": "recommend", "quote": "I now use Fable to create a PLAN.md.", "excerpt": "I have an 'ask' alias in my shell that just uses Haiku. I use it daily for pretty much everything For more long work, I now use Fable to create a PLAN.md. I tell it to make a plan that will be executed by other models, and most of the time it ends up choosing Opus or Sonnet. I didn't start doing this recently. Before that, I would just use the top model for everything. Splitting the work across different models depen", "context": "", "use_case": "planning long work", "switched_from": "", "story": "Ask HN: What default model do you use and why?"}]}, {"label": "Code quality", "summary": "Commenters say its code is more reliable and careful than other models', with fewer janky results.", "quotes": [{"id": 801351969, "item_id": 50084498, "url": "https://news.ycombinator.com/item?id=50084498", "by": "jwr", "date": "2026-10-14", "stance": "recommend", "quote": "With newer models, and especially large ones like Fable, it just doesn't happen anymore.", "excerpt": "> Every time I stop reviewing the code from Claude I later find janky crap I'm very curious about that because that has not been my experience at all. I used to find janky crap from older models and I had to review every line of code very carefully. With newer models, and especially large ones like Fable, it just doesn't happen anymore. The code that I get is what I would have written. Actually, it's usually better b", "context": "", "use_case": "AI coding", "switched_from": "", "story": "Redesigning source control at Anthropic"}, {"id": 794870257, "item_id": 49679391, "url": "https://news.ycombinator.com/item?id=49679391", "by": "qskousen", "date": "2026-09-13", "stance": "recommend", "quote": "Fable 5.1 was noticeably better at getting things right and doing things in a good reliable way.", "excerpt": "I have a pet project I have been working away on for some time that involves building GPU backends for various cards in Zig, lots of complex stuff in it. Lately I mostly use Opus 5, it can pretty reliably plug away at things but it does mess stuff up occasionally. For this codebase, Fable 5.1 was noticeably better at getting things right and doing things in a good reliable way. Of course, I can only use Fable for a b", "context": "", "use_case": "tough coding tasks", "switched_from": "", "story": "Everyone should slow down AI development except for me"}, {"id": 794786592, "item_id": 49674162, "url": "https://news.ycombinator.com/item?id=49674162", "by": "jamesponddotco", "date": "2026-09-12", "stance": "recommend", "quote": "Fable is the only one that fucking listens to me and writes proper Go code", "excerpt": "Given that Fable is the only one that fucking listens to me and writes proper Go code, that\u2019s what I use for everything code-related. At least on my personal projects. If I don\u2019t care about the code, i.e. I\u2019m writing something quick just to test something, I\u2019ll use GPT Astra to save my Fable tokens. For conversations and research I usually go with GPT Astra Pro. For Home Assistant I use a combination of Grok 4.20 and", "context": "", "use_case": "code", "switched_from": "", "story": "Ask HN: What default model do you use and why?"}]}, {"label": "Hard problems", "summary": "Used for a math proof, unsolved ciphers, a Python interpreter in Rust and computational biology.", "quotes": [{"id": 797576496, "item_id": 49848531, "url": "https://news.ycombinator.com/item?id=49848531", "by": "atemerev", "date": "2026-09-25", "stance": "recommend", "quote": "I used Fable 5.1 and Astra to solve and prove a long-standing mathematical problem", "excerpt": "Just recently I used Fable 5.1 and Astra to solve and prove a long-standing mathematical problem I was always interested in: the shoreline search problem (a ship in total fog is at unknown distance from the shore [infinite line], what is the optimal trajectory?) A particular kind of logarithmic spiral was conjectured 33 years ago; a few days ago, I have obtained the proof. How routine it has become. https://arxiv.org", "context": "mathematician", "use_case": "mathematical proofs", "switched_from": "", "story": "Yes, Claude can do nine loops"}, {"id": 795415840, "item_id": 49713490, "url": "https://news.ycombinator.com/item?id=49713490", "by": "Dbourdeau", "date": "2026-09-15", "stance": "recommend", "quote": "Mostly solved with Fable 5.1 and Opus 5.", "excerpt": "Following on others solving Ciphers with frontier LLMs, I wanted to give it a try. While I can't validate a new math proof, I can understand translated Ciphertext since its legible (numbers into readable text that makes sense). So far I've solved or made significant contributions to 6 'unsolved' ciphers, with more in my crosshairs. Mostly solved with Fable 5.1 and Opus 5. I've worked on a few 'famous' ones, seemingly", "context": "", "use_case": "solving ancient ciphers", "switched_from": "", "story": "Solving Ancient Ciphers with Claude"}, {"id": 803646848, "item_id": 50227928, "url": "https://news.ycombinator.com/item?id=50227928", "by": "genxy", "date": "2026-10-25", "stance": "recommend", "quote": "I used Fable 5.1 and Opus 5.5 to build a Python interpreter in Rust. Appears to pass all the 3.12 code I throw at it.", "excerpt": "I used Fable 5.1 and Opus 5.5 to build a Python interpreter in Rust. Appears to pass all the 3.12 code I throw at it. To get a toy version that can run most things takes under 2 hrs. Half of that time is collecting the tests you want to pass, how it should evaluate the code, what your criteria are, etc.", "context": "", "use_case": "building a Python interpreter", "switched_from": "", "story": "We should be more ambitious with the software we build"}]}, {"label": "Code reviews", "summary": "Commenters praise the reviews it gives across specs and codebases.", "quotes": [{"id": 795254544, "item_id": 49703409, "url": "https://news.ycombinator.com/item?id=49703409", "by": "criley2", "date": "2026-09-14", "stance": "recommend", "quote": "Fable 5.1 delivers incredible reviews.", "excerpt": "I have been writing an internal code review tool that is a bit maximalist. I created subagents for many internal domains and technologies we manage, with prompts focused on best practices, common problems, owasp guidelines, etc), a separate tier of wider band subagents (design, rollout, security/privacy), and a final agent at the top orchestrating and combining. I also use adversarial validator passes against all fin", "context": "works where privacy is enforced by law", "use_case": "internal code review", "switched_from": "", "story": "GPT-5.6 Luna vs. GPT-6 Astra: Is a $1.20 Model Good Enough for Code Review?"}, {"id": 794800036, "item_id": 49675002, "url": "https://news.ycombinator.com/item?id=49675002", "by": "jakevoytko", "date": "2026-09-12", "stance": "recommend", "quote": "I have Fable (Medium) review across the product spec, tech spec, and all the codebases", "excerpt": "My workhorse is Opus orchestrating Sonnet or Sol orchestrating Luna, depending on which ecosystem I'm using this week. I'm a backend engineer, so for tasks that either interact with the machine learning or client codebases, I have Fable (Medium) review across the product spec, tech spec, and all the codebases after the implementation is done (haven't tried this workflow with Astra yet, don't know if it's a good revie", "context": "backend engineer", "use_case": "code and spec review", "switched_from": "", "story": "Ask HN: What default model do you use and why?"}, {"id": 794875328, "item_id": 49679708, "url": "https://news.ycombinator.com/item?id=49679708", "by": "pdantix", "date": "2026-09-13", "stance": "recommend", "quote": "having fable plan and orchestrate with opus implemention + fable reviews, is my go-to.", "excerpt": "personally, i would not rely on opus 5 end to end as it'll start getting into walls of comment slop and shitting up the codebase similar to gpt 5.5's isRecord meme. on the other hand, having fable plan and orchestrate with opus implemention + fable reviews, is my go-to. if you give fable your guidelines up front or in your {claude,agents}.md, it will keep opus on a tight leash. opus can still write great code almost ", "context": "", "use_case": "planning and reviewing code", "switched_from": "", "story": "Real-SWE: Benchmarking AI models on private, real-world, enterprise codebases"}]}, {"label": "PCB design", "summary": "Some report good results designing and routing circuit boards with KiCad tooling.", "quotes": [{"id": 801902288, "item_id": 50118893, "url": "https://news.ycombinator.com/item?id=50118893", "by": "sgnelson", "date": "2026-10-16", "stance": "recommend", "quote": "I had Fable work on it, and it made a good board, after some trial and error.", "excerpt": "I've had good luck doing this recently with my own set of similar tools for a high altitude balloon tracker. The PCB has a transmitter module, pico, gps, barometer, accelerometer, magnetometer and extra flash plus all the power and rf filtering needs. It also needs to withstand ~ -40C temps. I also had size constraints. It did a good job of making compromises while designing a 4 layer board for prototyping. I had Fab", "context": "builds high altitude balloon trackers; helps with FabAcademy", "use_case": "PCB design", "switched_from": "", "story": "AI-assisted PCB design for KiCad 10"}, {"id": 801893728, "item_id": 50118358, "url": "https://news.ycombinator.com/item?id=50118358", "by": "ckemere", "date": "2026-10-16", "stance": "recommend", "quote": "I\u2019ve had really good results recently with Fable 5.1.", "excerpt": "I\u2019ve had really good results recently with Fable 5.1. The typical process that it uses for routing is to pin a few key traces (using the kicad python api) and then use freerouting for the others. I\u2019ve also noticed that if I push hard for \u201ccarefully designed\u201d routes, eg for a 38 pin connector, it will write code and make much better looking results. I\u2019d say that at the moment it still requires you to look at its resul", "context": "", "use_case": "PCB routing", "switched_from": "", "story": "AI-assisted PCB design for KiCad 10"}, {"id": 796289474, "item_id": 49768092, "url": "https://news.ycombinator.com/item?id=49768092", "by": "tonyarkles", "date": "2026-09-19", "stance": "recommend", "quote": "Fable took my Markdown notes, asked questions to fill in the blanks, and put together a plan.", "excerpt": "As a quick WIP update: - The first board, which I had already done manual schematic capture, floor planning, and some routing on (but then I lost steam and it\u2019s been sitting at 90% for 6 months) was finished quite readily by Opus + KiCadRoutingTools. - A second significantly more complex board that has been sitting at the \u201cI\u2019ve done a bunch of planning in Obsidian but haven\u2019t started schematic capture yet\u201d stage\u2026 Fab", "context": "", "use_case": "PCB planning", "switched_from": "", "story": "This PCB is brought to you by Fable 5"}]}], "caveats": [{"label": "Cost and limits", "summary": "Commenters say it is expensive and exhausts usage limits fast, sometimes on a separate meter.", "quotes": [{"id": 796647731, "item_id": 49790483, "url": "https://news.ycombinator.com/item?id=49790483", "by": "swalsh", "date": "2026-09-21", "stance": "warn", "quote": "Fable chewed through my usage too quickly to be practical", "excerpt": "Codex has become my goto tooling. I used to be a Claude Max subscriber, but I was becoming disappointed with the quality of the output from Opus 5. Fable chewed through my usage too quickly to be practical. Moving to a Pro account w/ Codex was a big improvement. Sol had great output, and the usage was more than sufficient for most of my needs. However astra does tend to chew up usage, so when i've done to much of tha", "context": "", "use_case": "", "switched_from": "", "story": "Grok 4.7"}, {"id": 796957489, "item_id": 49809843, "url": "https://news.ycombinator.com/item?id=49809843", "by": "matheusmoreira", "date": "2026-09-22", "stance": "warn", "quote": "Anthropic has a separate meter for Fable. I used to get like five Fable sessions per week and that's it.", "excerpt": "Anthropic has a separate meter for Fable. I used to get like five Fable sessions per week and that's it. OpenAI has no such nonsense. No separate meter. No five hour limits. I get to use Astra at max effort on literally every task if I want to, and even this somehow lasts me several days. Anthropic got caught playing stupid \"20x refers to the 5h limit\" word games with their customers. Meanwhile, I have statistically ", "context": "former Anthropic customer", "use_case": "coding", "switched_from": "", "story": "GPT-6 Sol and Luna"}, {"id": 794790496, "item_id": 49674406, "url": "https://news.ycombinator.com/item?id=49674406", "by": "montroser", "date": "2026-09-12", "stance": "warn", "quote": "In cost-per-task, it's a 32x multiple Fable vs DS41.", "excerpt": "It pains me to see so, so many dollars going to Anthropic. There is either massive inefficiency in there, or else its users are getting fleeced. In cost-per-task, it's a 32x multiple Fable vs DS41. Either it really does cost them more electricity/gpu, or else they are succeeding convincing the masses that Anthropic's models are necessary for everyday tasks, when this is truly just not the case. Either way seems a tra", "context": "", "use_case": "programming", "switched_from": "", "story": "Ask HN: What default model do you use and why?"}]}, {"label": "Writing style", "summary": "Some find its prose and word choice poor and hard to steer.", "quotes": [{"id": 804494592, "item_id": 50280912, "url": "https://news.ycombinator.com/item?id=50280912", "by": "n0on3", "date": "2026-10-28", "stance": "warn", "quote": "Fable 5.1 choice of words and general \u201cwords salad\u201d expression are horrific.", "excerpt": "Speaking for a user-facing use, not autonomous coding or agentic workflows: Fable 5.1 choice of words and general \u201cwords salad\u201d expression are horrific. It starts not so good, then as the context grows (still way below any reasonable rotting threshold) it gets completely insane. It\u2019s pointless for a model to be extremely capable if it gives you brain damage. I don\u2019t know if it\u2019s because of the watermark thingy, all I", "context": "", "use_case": "user-facing writing", "switched_from": "", "story": "Gemini 4 Carbon"}, {"id": 796864993, "item_id": 49804062, "url": "https://news.ycombinator.com/item?id=49804062", "by": "visarga", "date": "2026-09-22", "stance": "warn", "quote": "The poor writing style made even models like Fable look worse than Sonnet 3.7 to me.", "excerpt": "I tried a lot to get rid of Claude vibes and I failed. I made blacklists, I used prompt evolution to test against good human writing samples, even created a dynamic policy for dialogue, a cli tool used to steer agents dynamically by well chosen questions. The poor writing style made even models like Fable look worse than Sonnet 3.7 to me.", "context": "", "use_case": "writing", "switched_from": "", "story": "I don't want to read what you didn't write"}, {"id": 794870257, "item_id": 49679391, "url": "https://news.ycombinator.com/item?id=49679391", "by": "qskousen", "date": "2026-09-13", "stance": "recommend", "quote": "Fable 5.1 was noticeably better at getting things right and doing things in a good reliable way.", "excerpt": "I have a pet project I have been working away on for some time that involves building GPU backends for various cards in Zig, lots of complex stuff in it. Lately I mostly use Opus 5, it can pretty reliably plug away at things but it does mess stuff up occasionally. For this codebase, Fable 5.1 was noticeably better at getting things right and doing things in a good reliable way. Of course, I can only use Fable for a b", "context": "", "use_case": "tough coding tasks", "switched_from": "", "story": "Everyone should slow down AI development except for me"}]}, {"label": "Perceived nerfing", "summary": "A few commenters believe it got worse after release.", "quotes": [{"id": 804179971, "item_id": 50261248, "url": "https://news.ycombinator.com/item?id=50261248", "by": "Roark66", "date": "2026-10-27", "stance": "warn", "quote": "or fable before they lobotomised it", "excerpt": "Actually more AI agents do the work faster and cheaper. I use such orchestration for those 2 reasons. One, a supervisor model like Opus 5.5 (or fable before they lobotomised it) limited to 200k of context and told to keep it's progress updated in a text file is smart enough to understand the whole task. It farms smaller sub tasks to a fleet of 7 Qwen3.8-Flash-Next agents. Not only it uses up 10% of the tokens opus wo", "context": "", "use_case": "supervisor agent", "switched_from": "", "story": "The Mythical Man-Month in the Age of AI"}, {"id": 796649104, "item_id": 49790569, "url": "https://news.ycombinator.com/item?id=49790569", "by": "varispeed", "date": "2026-09-21", "stance": "warn", "quote": "I stopped using Fable long time ago. It's worse than Sonnet.", "excerpt": "I stopped using Fable long time ago. It's worse than Sonnet. Opus is not much better. This cycle of new model running at full quantisation and then nerfed few days / weeks after premiere should be called out. Anthropic should also drop the adaptive reasoning scam. If I pay for Fable, I should get full, not nerfed model at honest pricing. Regulators should investigate them. OpenAI is no different. Astra has basically ", "context": "", "use_case": "", "switched_from": "", "story": "Fable 5 \u2013 Median thinking declined in August"}, {"id": 795196929, "item_id": 49699808, "url": "https://news.ycombinator.com/item?id=49699808", "by": "varispeed", "date": "2026-09-14", "stance": "warn", "quote": "I have not tried Fable on it, because it was a car crash in previous similar task.", "excerpt": "The AI has gotten substantially worse. With Astra and Fable I am no longer able to finish tasks I used to be able to do and I revert to older models. Latest models go into reasoning loops, commit logic errors routinely and cannot even spot them by themselves and many more. For instance I have been trying to come up with generalised model of certain non-linear behaviour and had Astra working on it for 4 days. It produ", "context": "", "use_case": "", "switched_from": "", "story": "Fable 5.1 Solves the Cyphral Distich, a 370-year-old cipher"}]}, {"label": "Reasoning errors", "summary": "Some report it convincing itself of wrong conclusions or making basic mistakes.", "quotes": [{"id": 796718034, "item_id": 49794877, "url": "https://news.ycombinator.com/item?id=49794877", "by": "Sayrus", "date": "2026-09-21", "stance": "warn", "quote": "Opus and Fable always state something, then try to \"prove\" it but end up convincing themselves of the wrong thing.", "excerpt": "Same here. Moved from Opus to GLM 5.2 to 5.3 and I've been pretty happy with the result. Mainly, it doesn't hallucinate and convince itself of mistake so it's good at retrieving information or asking the user for it. Opus and Fable always state something, then try to \"prove\" it but end up convincing themselves of the wrong thing. Having subagents for retrieval and validation helped but were not enough.", "context": "", "use_case": "coding", "switched_from": "", "story": "Fable 5 \u2013 Median thinking declined in August"}, {"id": 795568720, "item_id": 49723045, "url": "https://news.ycombinator.com/item?id=49723045", "by": "melvinroest", "date": "2026-09-16", "stance": "warn", "quote": "LLMs can be unbelievably stupid. It recently happened a few times with Fable 5.1 as well.", "excerpt": "Yea I get the bearishness from my own personal experience. Personally, I use LLMs for a lot of things. Oftentimes, I'm a think out loud type of person so even having something that feels like a rubber duck, but more competent, is already amazing for me. And LLMs are a lot more competent than a rubber duck. But especially sometimes I've noticed that LLMs can be unbelievably stupid. It recently happened a few times wit", "context": "", "use_case": "", "switched_from": "", "story": "Why I'm still bearish on LLMs after Navier-Stokes"}, {"id": 798993312, "item_id": 49937082, "url": "https://news.ycombinator.com/item?id=49937082", "by": "rancar2", "date": "2026-10-02", "stance": "warn", "quote": "there\u2019s concerning behavior for Fable on this front", "excerpt": "Doctors are good with getting a shorthand note they can scan prior to appointment about updates, etc reverent to the appointment that day. My experience is that AI models even with expert curations cannot currently do well by itself and would waste a doctor\u2019s time and could be misleading too. As such as noted by the parent comment, that note would need to be summary sheet that AI touches would have to list facts and ", "context": "", "use_case": "medical AI", "switched_from": "", "story": "Show HN: We built a real-time desktop companion for patient's telehealth visits"}]}, {"label": "Unsupervised drift", "summary": "Commenters say it can over-engineer and cannot keep a codebase organized on long unsupervised tasks.", "quotes": [{"id": 794825491, "item_id": 49676593, "url": "https://news.ycombinator.com/item?id=49676593", "by": "karmakaze", "date": "2026-09-12", "stance": "warn", "quote": "Fable can venture into over-engineering itself into unnecessary complications.", "excerpt": "Personally I'm using Qwen3.8-27B (MXFP4 quant W4A8) locally hosted on a pair of AMD GPUs (with DeepSeek Harness). It starts at 250 tokens/sec down to 120 past 128k context. At work mostly Opus 4.8 (sometimes a GPT or Gemini 3.1 Pro). I find Opus 5 chatty/slower and Fable can venture into over-engineering itself into unnecessary complications.", "context": "", "use_case": "", "switched_from": "", "story": "Ask HN: What default model do you use and why?"}, {"id": 796904896, "item_id": 49806556, "url": "https://news.ycombinator.com/item?id=49806556", "by": "physicallyIllfr", "date": "2026-09-22", "stance": "warn", "quote": "not even fable can keep a codebase organized on long running unspervised tasks", "excerpt": "No its not I use llms just as much as the next guy and not even fable can keep a codebase organized on long running unspervised tasks.", "context": "", "use_case": "long-running coding tasks", "switched_from": "", "story": "GPT-6 Sol and Luna"}]}], "who": [{"by": "boogiefromzk", "context": "built an agentic delegation scheme", "url": "https://news.ycombinator.com/item?id=50251451", "date": "2026-10-26"}, {"by": "sgnelson", "context": "builds high altitude balloon trackers; helps with FabAcademy", "url": "https://news.ycombinator.com/item?id=50118893", "date": "2026-10-16"}, {"by": "laurels-marts", "context": "uses Codex at work", "url": "https://news.ycombinator.com/item?id=49905209", "date": "2026-09-30"}, {"by": "ch4s3", "context": "building an iOS app", "url": "https://news.ycombinator.com/item?id=49903518", "date": "2026-09-30"}, {"by": "atemerev", "context": "mathematician", "url": "https://news.ycombinator.com/item?id=49848531", "date": "2026-09-25"}, {"by": "jacobgold", "context": "pays $200/mo Claude subscription", "url": "https://news.ycombinator.com/item?id=49804117", "date": "2026-09-22"}, {"by": "SyneRyder", "context": "built voice harness in Go", "url": "https://news.ycombinator.com/item?id=49731152", "date": "2026-09-16"}, {"by": "zopf", "context": "startup CTO returning to trading systems", "url": "https://news.ycombinator.com/item?id=49705167", "date": "2026-09-14"}, {"by": "criley2", "context": "works where privacy is enforced by law", "url": "https://news.ycombinator.com/item?id=49703409", "date": "2026-09-14"}, {"by": "arjie", "context": "has a personal agent on local hardware", "url": "https://news.ycombinator.com/item?id=49700364", "date": "2026-09-14"}, {"by": "rajay99", "context": "computational biology", "url": "https://news.ycombinator.com/item?id=49691117", "date": "2026-09-14"}, {"by": "ddahlen", "context": "not a frontend developer", "url": "https://news.ycombinator.com/item?id=49690394", "date": "2026-09-14"}], "tags": ["for agent orchestration", "for code review", "for PCB design", "for research and math", "for Claude Max subscribers"], "use_cases": ["coding", "code review", "coding agent", "planning", "work coding", "ai coding"], "compared": [{"slug": "claude-opus", "name": "Claude Opus", "co": 239, "switched_from": 15, "switched_to": 5}, {"slug": "gpt-sol", "name": "GPT Sol", "co": 56, "switched_from": 1, "switched_to": 5}, {"slug": "codex", "name": "Codex", "co": 54, "switched_from": 0, "switched_to": 3}, {"slug": "claude-code", "name": "Claude Code", "co": 50, "switched_from": 0, "switched_to": 0}, {"slug": "gpt-5", "name": "GPT-5", "co": 39, "switched_from": 1, "switched_to": 2}, {"slug": "claude", "name": "Claude", "co": 40, "switched_from": 0, "switched_to": 0}, {"slug": "gpt-astra", "name": "GPT Astra", "co": 31, "switched_from": 1, "switched_to": 2}, {"slug": "claude-sonnet", "name": "Claude Sonnet", "co": 33, "switched_from": 0, "switched_to": 0}], "makers": [{"id": 792412944, "item_id": 49525809, "url": "https://news.ycombinator.com/item?id=49525809", "by": "felixrieseberg", "date": "2026-09-01", "stance": "recommend", "quote": "I think Fable 5.1 is a big improvement in writing style.", "excerpt": "(I work at Anthropic) Beyond all the benchmarks, I think Fable 5.1 is a big improvement in writing style. It sounds a lot less stereotypically like other Claude models, has (imho) a much more natural style, and responds to my style instructions more reliably. More work to be done (and we will!) but reading better prose makes me so much happier. Another point I expect not to get much attention until it all happens at ", "context": "works at Anthropic", "use_case": "", "switched_from": "", "story": "Claude Fable 5.1 and Claude Mythos 5.1"}], "threads": [{"url": "https://news.ycombinator.com/item?id=48463808", "title": "Claude Fable 5", "mentions": 35}, {"url": "https://news.ycombinator.com/item?id=49525378", "title": "Claude Fable 5.1 and Claude Mythos 5.1", "mentions": 19}, {"url": "https://news.ycombinator.com/item?id=49672966", "title": "Ask HN: What default model do you use and why?", "mentions": 11}, {"url": "https://news.ycombinator.com/item?id=48849066", "title": "GPT-5.6", "mentions": 11}, {"url": "https://news.ycombinator.com/item?id=48511072", "title": "US Government directive to suspend access to Fable 5 and Mythos 5", "mentions": 9}, {"url": "https://news.ycombinator.com/item?id=49803892", "title": "Claude Opus 5.5", "mentions": 7}], "sentiment_by_month": {"2025-11": {}, "2025-12": {}, "2026-01": {}, "2026-02": {}, "2026-03": {"recommend": 1}, "2026-04": {}, "2026-05": {}, "2026-06": {"warn": 34, "neutral": 16, "recommend": 44, "compare": 6}, "2026-07": {"recommend": 56, "warn": 25, "neutral": 33, "compare": 12}, "2026-08": {"compare": 9, "recommend": 26, "neutral": 13, "warn": 16}, "2026-09": {"neutral": 24, "recommend": 44, "warn": 21, "compare": 7}, "2026-10": {"recommend": 11, "warn": 4, "neutral": 3, "compare": 2}}, "quotes": [{"id": 804750817, "item_id": 50296926, "url": "https://news.ycombinator.com/item?id=50296926", "by": "jrockway", "date": "2026-10-29", "stance": "recommend", "quote": "Fable is better at playing along with my \"You are a dog\" instruction in CLAUDE.md.", "excerpt": "Opus 5.5 is definitely good. I use both all day and it's really hard to say which is which from looking at the outputs; I can't read a PR and say \"this was Fable\" or \"this was Opus\" like I could pre-5.5. Fable is better at playing along with my \"You are a dog\" instruction in CLAUDE.md. It is willing to be fun and playful, whereas Opus just wants to finish the job and go home. Friend vs. coworker, basically. For examp", "context": "", "use_case": "coding", "switched_from": "", "story": "Ask HN: Opus 5.5 suddenly dumb as rocks?"}, {"id": 804634402, "item_id": 50289650, "url": "https://news.ycombinator.com/item?id=50289650", "by": "dichiara19", "date": "2026-10-29", "stance": "compare", "quote": "I don\u2019t think it measures up to Fable 5.1 or Astra", "excerpt": "I want to talk about my experience with Argon, noting that I used it via OpenRouter on a Raspberry Pi, spending around $15 in credit. My experience is that Google is aiming to create a generalist model \u2013 but a truly generalist one! In fact, I find it particularly accurate in general terms and even in long conversations, maintaining excellent consistency and a wealth of information in response to complex questions. I\u2019", "context": "", "use_case": "coding agent", "switched_from": "", "story": "Gemini 4 Carbon is the new #1 on the Artificial Analysis Intelligence Index v5"}, {"id": 804494592, "item_id": 50280912, "url": "https://news.ycombinator.com/item?id=50280912", "by": "n0on3", "date": "2026-10-28", "stance": "warn", "quote": "Fable 5.1 choice of words and general \u201cwords salad\u201d expression are horrific.", "excerpt": "Speaking for a user-facing use, not autonomous coding or agentic workflows: Fable 5.1 choice of words and general \u201cwords salad\u201d expression are horrific. It starts not so good, then as the context grows (still way below any reasonable rotting threshold) it gets completely insane. It\u2019s pointless for a model to be extremely capable if it gives you brain damage. I don\u2019t know if it\u2019s because of the watermark thingy, all I", "context": "", "use_case": "user-facing writing", "switched_from": "", "story": "Gemini 4 Carbon"}, {"id": 804471041, "item_id": 50279440, "url": "https://news.ycombinator.com/item?id=50279440", "by": "tern", "date": "2026-10-28", "stance": "compare", "quote": "I only use Fable when \"knowing about more stuff\" is important (like research, exploring ideas)", "excerpt": "I only use Fable when \"knowing about more stuff\" is important (like research, exploring ideas), but I find Opus 5.5 writes and expresses itself more clearly and is reliable for anything I throw at it, so I use it much more. It's possible Fable is better, but since it word salads more quickly, it's hard to tell.", "context": "", "use_case": "research, exploring ideas", "switched_from": "", "story": "Gemini 4 Carbon"}, {"id": 804179971, "item_id": 50261248, "url": "https://news.ycombinator.com/item?id=50261248", "by": "Roark66", "date": "2026-10-27", "stance": "warn", "quote": "or fable before they lobotomised it", "excerpt": "Actually more AI agents do the work faster and cheaper. I use such orchestration for those 2 reasons. One, a supervisor model like Opus 5.5 (or fable before they lobotomised it) limited to 200k of context and told to keep it's progress updated in a text file is smart enough to understand the whole task. It farms smaller sub tasks to a fleet of 7 Qwen3.8-Flash-Next agents. Not only it uses up 10% of the tokens opus wo", "context": "", "use_case": "supervisor agent", "switched_from": "", "story": "The Mythical Man-Month in the Age of AI"}, {"id": 804023216, "item_id": 50251451, "url": "https://news.ycombinator.com/item?id=50251451", "by": "boogiefromzk", "date": "2026-10-26", "stance": "recommend", "quote": "where I use Fable and its limits spent slower than overall account limit", "excerpt": "Last months I was searching for a way to make smart AI model submit and verify the work to cheap model so that it will cost less, and after many try and failures I was able to get stable result where I use Fable and its limits spent slower than overall account limit because most context intensive work is done by Sonnet. At the same time the response quaility is still Fable. I also made sure Opus controls Grok in Curs", "context": "built an agentic delegation scheme", "use_case": "orchestrating agentic delegation", "switched_from": "", "story": "Cost-effective agentic delegation scheme for Claude Code, Cursor and Codex"}, {"id": 803646848, "item_id": 50227928, "url": "https://news.ycombinator.com/item?id=50227928", "by": "genxy", "date": "2026-10-25", "stance": "recommend", "quote": "I used Fable 5.1 and Opus 5.5 to build a Python interpreter in Rust. Appears to pass all the 3.12 code I throw at it.", "excerpt": "I used Fable 5.1 and Opus 5.5 to build a Python interpreter in Rust. Appears to pass all the 3.12 code I throw at it. To get a toy version that can run most things takes under 2 hrs. Half of that time is collecting the tests you want to pass, how it should evaluate the code, what your criteria are, etc.", "context": "", "use_case": "building a Python interpreter", "switched_from": "", "story": "We should be more ambitious with the software we build"}, {"id": 803445841, "item_id": 50215365, "url": "https://news.ycombinator.com/item?id=50215365", "by": "3princip", "date": "2026-10-23", "stance": "recommend", "quote": "(max plan, haven't run out of the hourly or weekly limit once yet", "excerpt": "Same plan, just asked it to review a repo that Fable has been helping out with (max plan, haven't run out of the hourly or weekly limit once yet, and I use the same sub on multiple projects at once). Kimi didn't even finish the scan, ran out of the 5hr budget in less than 10 minutes. Unusable, unfortunately.", "context": "", "use_case": "helping on repos", "switched_from": "", "story": "Kimi K3.1 Pro and K3.1 Pro Highspeed"}, {"id": 802787504, "item_id": 50174219, "url": "https://news.ycombinator.com/item?id=50174219", "by": "dhorthy", "date": "2026-10-20", "stance": "neutral", "quote": "fable scored 33%, while on the full challenge it landed closer to 14-15% range", "excerpt": "Yeah and I started with subsets although I did not find them representative, eg I picked 6 challenges for a subset and fable scored 33%, while on the full challenge it landed closer to 14-15% range The part II post linked in my top comment has a list of the challenges I used for that. To run them, just point your model at the paper, it can find the repo, etc and with a key or sub, get it running", "context": "", "use_case": "benchmark subset", "switched_from": "", "story": "SlopCodeBench 10/2026 \u2013 Opus 5.5, Astra wins overall, xhigh considered harmful"}, {"id": 802763585, "item_id": 50172724, "url": "https://news.ycombinator.com/item?id=50172724", "by": "CharlieDigital", "date": "2026-10-20", "stance": "recommend", "quote": "\"Distinguished Member of Technical Staff\" -> Fable, plan mode.", "excerpt": "I use the frontier models only in plan mode; coding is held out to cheaper models. I run Paseo as an orchestrator with agent profiles. \"Distinguished Member of Technical Staff\" -> Fable, plan mode. \"End-User Customer Advocate\" -> Astra, plan mode. A few others. Basically, these agents (with profile prompts) can only be \"consulted\" for input on design elements; it's OK if they produce a bit of verbosity here. This has", "context": "", "use_case": "planning/design", "switched_from": "", "story": "SlopCodeBench 10/2026 \u2013 Opus 5.5, Astra wins overall, xhigh considered harmful"}, {"id": 801902288, "item_id": 50118893, "url": "https://news.ycombinator.com/item?id=50118893", "by": "sgnelson", "date": "2026-10-16", "stance": "recommend", "quote": "I had Fable work on it, and it made a good board, after some trial and error.", "excerpt": "I've had good luck doing this recently with my own set of similar tools for a high altitude balloon tracker. The PCB has a transmitter module, pico, gps, barometer, accelerometer, magnetometer and extra flash plus all the power and rf filtering needs. It also needs to withstand ~ -40C temps. I also had size constraints. It did a good job of making compromises while designing a 4 layer board for prototyping. I had Fab", "context": "builds high altitude balloon trackers; helps with FabAcademy", "use_case": "PCB design", "switched_from": "", "story": "AI-assisted PCB design for KiCad 10"}, {"id": 801893728, "item_id": 50118358, "url": "https://news.ycombinator.com/item?id=50118358", "by": "ckemere", "date": "2026-10-16", "stance": "recommend", "quote": "I\u2019ve had really good results recently with Fable 5.1.", "excerpt": "I\u2019ve had really good results recently with Fable 5.1. The typical process that it uses for routing is to pin a few key traces (using the kicad python api) and then use freerouting for the others. I\u2019ve also noticed that if I push hard for \u201ccarefully designed\u201d routes, eg for a 38 pin connector, it will write code and make much better looking results. I\u2019d say that at the moment it still requires you to look at its resul", "context": "", "use_case": "PCB routing", "switched_from": "", "story": "AI-assisted PCB design for KiCad 10"}, {"id": 801738320, "item_id": 50108645, "url": "https://news.ycombinator.com/item?id=50108645", "by": "HamadMalikKhan", "date": "2026-10-15", "stance": "recommend", "quote": "With the new Fable 5.1, even the how part is getting replaced.", "excerpt": "I wrote under 5% of the code by hand. It was mainly telling the AI to build x and how to build it. With the new Fable 5.1, even the how part is getting replaced. For this project, the maintaince is not too big of a pain, but we are working on another project where it's an absolute pain. Every PR is like 100k lines change, simple bug fix needs 600k tokens. We are probably doing a choppy job here. The size of the codeb", "context": "", "use_case": "AI code generation", "switched_from": "", "story": "Ask HN: Has anyone made real money from AI gen vibe coded app?"}, {"id": 801367472, "item_id": 50085467, "url": "https://news.ycombinator.com/item?id=50085467", "by": "menaerus", "date": "2026-10-14", "stance": "neutral", "quote": "their new model B (Fable) capabilities", "excerpt": "Well, Anthropic used model A to output the version 1, and then used model B to rewrite the version 1 into version 2, so the handoff was machine to machine and not human to machine. I read that part of an article as if it was their internal assessment test of their new model B (Fable) capabilities so that they can see if the new model can find interesting things but by intentionally being vague and without implementin", "context": "", "use_case": "rewriting code", "switched_from": "", "story": "Redesigning source control at Anthropic"}, {"id": 801351969, "item_id": 50084498, "url": "https://news.ycombinator.com/item?id=50084498", "by": "jwr", "date": "2026-10-14", "stance": "recommend", "quote": "With newer models, and especially large ones like Fable, it just doesn't happen anymore.", "excerpt": "> Every time I stop reviewing the code from Claude I later find janky crap I'm very curious about that because that has not been my experience at all. I used to find janky crap from older models and I had to review every line of code very carefully. With newer models, and especially large ones like Fable, it just doesn't happen anymore. The code that I get is what I would have written. Actually, it's usually better b", "context": "", "use_case": "AI coding", "switched_from": "", "story": "Redesigning source control at Anthropic"}, {"id": 800403125, "item_id": 50025195, "url": "https://news.ycombinator.com/item?id=50025195", "by": "aftbit", "date": "2026-10-09", "stance": "neutral", "quote": "occasional jumps up to Astra or Fable", "excerpt": "I have had good luck rotating between the three tiers of GPT 5.6 with occasional jumps up to Astra or Fable. Most of my work is with GPT-5.6-Sol. Simple tasks like data extraction or trivial refactors (rename this variable etc) I often push down to Terra or Luna, or even to self-hosted Gemma4:31B. Very tricky stuff, like planning a new feature, design review, or code review of a complex change across multiple reposit", "context": "", "use_case": "harder tasks", "switched_from": "", "story": "Why isn't the industry freaking out about DeepSeek 4.1 Flash?"}, {"id": 800250992, "item_id": 50015687, "url": "https://news.ycombinator.com/item?id=50015687", "by": "yuhmahp", "date": "2026-10-09", "stance": "recommend", "quote": "We use Claude Fable to plan, and DeepSeek 4.1 Flash (hosted on DeepInfra) for everything else. Very cost effective.", "excerpt": "We use Claude Fable to plan, and DeepSeek 4.1 Flash (hosted on DeepInfra) for everything else. Very cost effective.", "context": "", "use_case": "planning", "switched_from": "", "story": "Why isn't the industry freaking out about DeepSeek 4.1 Flash?"}, {"id": 799427748, "item_id": 49964234, "url": "https://news.ycombinator.com/item?id=49964234", "by": "Livia_Zaharia", "date": "2026-10-05", "stance": "warn", "quote": "We switched to Fable but it was a bit dissapointing", "excerpt": "Hello, I am Livia Zaharia, officially an architect, also a parametric jewelery designer, machine learning enthusiast and a startup founder. What I wanted to share is the link above, where, to promote my pieces in a more interactive way I decided to try building from scratch a 3d version of a town where some pieces become buildings. I developed it together with my boyfriend being bord from waiting for our ML training ", "context": "architect, jewelry designer, startup founder", "use_case": "3D world generation", "switched_from": "GPT Astra", "story": null}, {"id": 798993312, "item_id": 49937082, "url": "https://news.ycombinator.com/item?id=49937082", "by": "rancar2", "date": "2026-10-02", "stance": "warn", "quote": "there\u2019s concerning behavior for Fable on this front", "excerpt": "Doctors are good with getting a shorthand note they can scan prior to appointment about updates, etc reverent to the appointment that day. My experience is that AI models even with expert curations cannot currently do well by itself and would waste a doctor\u2019s time and could be misleading too. As such as noted by the parent comment, that note would need to be summary sheet that AI touches would have to list facts and ", "context": "", "use_case": "medical AI", "switched_from": "", "story": "Show HN: We built a real-time desktop companion for patient's telehealth visits"}, {"id": 798881459, "item_id": 49930091, "url": "https://news.ycombinator.com/item?id=49930091", "by": "nl", "date": "2026-10-02", "stance": "recommend", "quote": "Fable is great.", "excerpt": "Having used them heavily for 3D model understanding since February I can say it's nuanced. Opus 4.6 and 4.7 were bad, but GPT 5.2 and above were very usable. Opus 4.8 was usable, but the GPT 5.x series was better. Fable is great. Opus 5.0 was interesting. It could solve some problems that Sol 5.x couldn't solve (applying a G2 curve on a 3 way corner where one face was a Bezier curve) but you had to be super prescript", "context": "", "use_case": "3D model understanding", "switched_from": "", "story": "Using Opus 5.5 to discover a new eyewitness record of the dodo"}, {"id": 798621667, "item_id": 49913854, "url": "https://news.ycombinator.com/item?id=49913854", "by": "spankalee", "date": "2026-09-30", "stance": "neutral", "quote": "I use a mix of Fable 5.1, Opus 5.5, and Gemini 3.8 Flash", "excerpt": "3.8 Flash is just quite good, and so is the Antigravity harness. I use a mix of Fable 5.1, Opus 5.5, and Gemini 3.8 Flash and Gemini holds it's own. Especially in writing, frontend, and sysadmin work. agy for configuring a NixOS system has been truly incredible.", "context": "", "use_case": "", "switched_from": "", "story": "Gemini 4 Argon"}, {"id": 798555745, "item_id": 49909734, "url": "https://news.ycombinator.com/item?id=49909734", "by": "mattdesl", "date": "2026-09-30", "stance": "neutral", "quote": "based on a formulation that Fable 5 originally proposed to me during a directed search", "excerpt": "I've also been working on a GPU curve renderer, Windfoil, based on a formulation that Fable 5 originally proposed to me during a directed search [1]. It is similar in some ways to Slug, not always as fast, but uses less shader storage (single band instead of two) and produces higher quality anti-aliasing i.e. closer to a box-filtered ground truth. It may be of interest to some game/graphics devs here... [1] https://g", "context": "", "use_case": "proposing algorithms", "switched_from": "", "story": "SDF vs. MSDF vs. Slug: GPU Text Rendering"}, {"id": 798483344, "item_id": 49905209, "url": "https://news.ycombinator.com/item?id=49905209", "by": "laurels-marts", "date": "2026-09-30", "stance": "recommend", "quote": "I tried out fable 5.1 the day it was released and coming from gpt-5.6-sol I was truly mind blown", "excerpt": "100% in agreement. I pay for OAI sub and also use Codex exclusively at work for the past 8 months. I tried out fable 5.1 the day it was released and coming from gpt-5.6-sol I was truly mind blown (both in terms of code and prose it was generating - outputs I could finally enjoy reading and looking at). Then when opus 5.5 came out, again same thing + far cheaper and faster. I went from using OAI exclusively the entire", "context": "uses Codex at work", "use_case": "coding and prose generation", "switched_from": "GPT-5.6 Sol", "story": "GPT 6.1 Sol: Near-Astra intelligence for a fifth of the price"}, {"id": 798456289, "item_id": 49903518, "url": "https://news.ycombinator.com/item?id=49903518", "by": "ch4s3", "date": "2026-09-30", "stance": "neutral", "quote": "fable was released and it seems my taste lines up with fable\u2019s training quite a bit", "excerpt": "It\u2019s funny, this is exactly how I used opus to design an iOS app I\u2019m building. Unfortunately fable was released and it seems my taste lines up with fable\u2019s training quite a bit[1]. Now my app looks like ai anyway. You win some you lose some I guess. [1] https://onda-cast.github.io/onda/", "context": "building an iOS app", "use_case": "app design", "switched_from": "", "story": "How our vibe coded website looks like a designer made it"}, {"id": 798435232, "item_id": 49902202, "url": "https://news.ycombinator.com/item?id=49902202", "by": "wslh", "date": "2026-09-29", "stance": "neutral", "quote": "I didn't experience issues when I used Fable via OpenRouter", "excerpt": "> > They are calling out specific providers who release powerful models without safeguards. > Good thing Fable refuses to answer my question about how children inherit blue eyes, it was the only piece of information I needed to finish my blue-eye super-bioweapon. FWIW: I didn't experience issues when I used Fable via OpenRouter checking for security issues in code but experienced them via the Claude desktop app.", "context": "", "use_case": "security review of code", "switched_from": "", "story": "GLM-5.3 and the spread of advanced cyber capabilities"}, {"id": 798398946, "item_id": 49899934, "url": "https://news.ycombinator.com/item?id=49899934", "by": "sanex", "date": "2026-09-29", "stance": "neutral", "quote": "he delegates complicated things to a project lead which is usually fable", "excerpt": "Yeah personally I use 3 levels of agents. My main who is either a sonnet or opus, he delegates complicated things to a project lead which is usually fable, and then he delegates everything to the dumbest possible model for the task.", "context": "", "use_case": "project lead agent for complex tasks", "switched_from": "", "story": "Dots: Always-on agents"}, {"id": 798211904, "item_id": 49888244, "url": "https://news.ycombinator.com/item?id=49888244", "by": "selcuka", "date": "2026-09-29", "stance": "neutral", "quote": "even though I use Fable for making plans", "excerpt": "I do that and can use up the $20 subscription, but I haven't been able to hit the limits with the 5x subscription so far (even though I use Fable for making plans).", "context": "", "use_case": "making plans", "switched_from": "", "story": "Sonnet 5.5"}, {"id": 797686674, "item_id": 49855417, "url": "https://news.ycombinator.com/item?id=49855417", "by": "LUmBULtERA", "date": "2026-09-26", "stance": "compare", "quote": "If I were to switch to Anthropic, Opus 5.5 or Fable 5.1 on high or better.", "excerpt": "Astra on Extra High. Knowledge work, the right answer is much more important than a fast or cheap answer. If I were to switch to Anthropic, Opus 5.5 or Fable 5.1 on high or better.", "context": "", "use_case": "", "switched_from": "", "story": "Ask HN: Which model do you use for work?"}, {"id": 797641777, "item_id": 49852611, "url": "https://news.ycombinator.com/item?id=49852611", "by": "themanmaran", "date": "2026-09-26", "stance": "recommend", "quote": "I use plan mode constantly (with opus/fable)", "excerpt": "I don't think me giving the model bad instructions is something a smarter model can solve. I use plan mode constantly (with opus/fable), and at least once a day I'll say something too vague or just dumb and it will sketch out the \"wrong\" solution in it's plan. Which is fine because it just put together a plan and didn't spend 10 minutes rearchitecting everything.", "context": "", "use_case": "planning and coding", "switched_from": "", "story": "Plan mode is dead"}, {"id": 797631632, "item_id": 49851977, "url": "https://news.ycombinator.com/item?id=49851977", "by": "nijave", "date": "2026-09-26", "stance": "recommend", "quote": "that was a full 180 back to coherency", "excerpt": "Personally, a combination of low tech and semi scientific tests based on what you normally do (redo the same task you used an older model for with a newer one). So far, Simon Willison's pelican bike \"benchmark\" is the only one I've found that shows Fable 5.1 beating Opus 5.5. My personal experience has been Opus is unusable on design work it's so terrible. Evaluating whether we should consolidate AWS DMS tasks (Postg", "context": "", "use_case": "research and design work", "switched_from": "Opus 5.5", "story": "Tokens too cheap to meter"}, {"id": 797619888, "item_id": 49851243, "url": "https://news.ycombinator.com/item?id=49851243", "by": "status_quo69", "date": "2026-09-25", "stance": "recommend", "quote": "I use fable 5.1", "excerpt": "I use fable 5.1 (tried opus but it lied to me 3 times in quick succession and ignored me in another) This is weird to ask because I feel like of course the model isn't omniscient? Isn't the whole point of iterating on a plan to assess impact, risk, know your (the user) variables, user impact, product impact, etc for making a change? I cannot count the times even in the past few months where I start a conversation wit", "context": "", "use_case": "", "switched_from": "Opus", "story": "Plan mode is dead"}, {"id": 797576496, "item_id": 49848531, "url": "https://news.ycombinator.com/item?id=49848531", "by": "atemerev", "date": "2026-09-25", "stance": "recommend", "quote": "I used Fable 5.1 and Astra to solve and prove a long-standing mathematical problem", "excerpt": "Just recently I used Fable 5.1 and Astra to solve and prove a long-standing mathematical problem I was always interested in: the shoreline search problem (a ship in total fog is at unknown distance from the shore [infinite line], what is the optimal trajectory?) A particular kind of logarithmic spiral was conjectured 33 years ago; a few days ago, I have obtained the proof. How routine it has become. https://arxiv.org", "context": "mathematician", "use_case": "mathematical proofs", "switched_from": "", "story": "Yes, Claude can do nine loops"}, {"id": 797575970, "item_id": 49848498, "url": "https://news.ycombinator.com/item?id=49848498", "by": "trevorkiw", "date": "2026-09-25", "stance": "compare", "quote": "i'm well aware of opus and fable's superior engineering qualities for writing code", "excerpt": "gpt 5.6 sol. assuming the prose is similar with the gpt 6 series, if it was available at work that's probably what i'd be using instead. i'm well aware of opus and fable's superior engineering qualities for writing code, however i find that as a swe who needs to understand every engineering decision that is made and is held accountable for their actions for safety critical code, understanding trumps speed every time.", "context": "SWE on safety critical code", "use_case": "writing code", "switched_from": "", "story": "Ask HN: Which model do you use for work?"}, {"id": 797321458, "item_id": 49832591, "url": "https://news.ycombinator.com/item?id=49832591", "by": "Aurornis", "date": "2026-09-24", "stance": "recommend", "quote": "but with Opus, Astra, or Fable it\u2019s great", "excerpt": "I\u2019ve found LLMs to be very good at teaching. Better than I expected. You have to tell it what you want out of the session. A kid could just type in the question and get the answer, but that\u2019s not tutoring. When I use them for learning I tell it my goal, summarize my current status, and then tell it to teach me incrementally with a quiz after each step. You can get fancy and have it reach back into history to throw in", "context": "", "use_case": "learning and tutoring", "switched_from": "", "story": "Tutoring company tells parents to save their money and 'use AI instead'"}, {"id": 796984273, "item_id": 49811517, "url": "https://news.ycombinator.com/item?id=49811517", "by": "jauntywundrkind", "date": "2026-09-23", "stance": "neutral", "quote": "I used Fable some a bit back", "excerpt": "I used Fable some a bit back but have been using mostly Codex, and it seems pretty clear to me: the big models are not there to do terminal bench. They're there to plan and strategize and tell lesser models what they should do in terminal bench. We've had a year of nearly every model getting quite good at programming. I think with Fable & Astra we are seeing models trained to think at a different level, and I'm not a", "context": "", "use_case": "coding", "switched_from": "", "story": "Claude Opus 5.5"}, {"id": 796957489, "item_id": 49809843, "url": "https://news.ycombinator.com/item?id=49809843", "by": "matheusmoreira", "date": "2026-09-22", "stance": "warn", "quote": "Anthropic has a separate meter for Fable. I used to get like five Fable sessions per week and that's it.", "excerpt": "Anthropic has a separate meter for Fable. I used to get like five Fable sessions per week and that's it. OpenAI has no such nonsense. No separate meter. No five hour limits. I get to use Astra at max effort on literally every task if I want to, and even this somehow lasts me several days. Anthropic got caught playing stupid \"20x refers to the 5h limit\" word games with their customers. Meanwhile, I have statistically ", "context": "former Anthropic customer", "use_case": "coding", "switched_from": "", "story": "GPT-6 Sol and Luna"}, {"id": 796955810, "item_id": 49809738, "url": "https://news.ycombinator.com/item?id=49809738", "by": "schipperai", "date": "2026-09-22", "stance": "recommend", "quote": "Fable is on a class of its own when it comes to coding and orchestration", "excerpt": "I have used Fable, Astra, Opus 5 and Sol 5.6 in Claude Code and Codex heavily since each of these have been released. Recently, I settled for Sol 5.6 as my daily driver, but with this release Opus 5.5 might be the one. Fable is on a class of its own when it comes to coding and orchestration, but it runs out pretty quick and is prohibitively expensive and slow. For me, Astra wasn't as good of an upgrade from Sol when ", "context": "", "use_case": "coding and orchestration", "switched_from": "", "story": "Claude Opus 5.5"}, {"id": 796904896, "item_id": 49806556, "url": "https://news.ycombinator.com/item?id=49806556", "by": "physicallyIllfr", "date": "2026-09-22", "stance": "warn", "quote": "not even fable can keep a codebase organized on long running unspervised tasks", "excerpt": "No its not I use llms just as much as the next guy and not even fable can keep a codebase organized on long running unspervised tasks.", "context": "", "use_case": "long-running coding tasks", "switched_from": "", "story": "GPT-6 Sol and Luna"}, {"id": 796896816, "item_id": 49806051, "url": "https://news.ycombinator.com/item?id=49806051", "by": "hombre_fatal", "date": "2026-09-22", "stance": "recommend", "quote": "My main process is to use Fable to reason and then spawn Opus subagents, and I get amazing results", "excerpt": "I mainly use Codex/Sol to review my plans drafted by Fable. But beyond that, Astra blows through usage limits too fast to be a daily driver and writes weird code despite what my \"house style\" is, and Codex is behind Claude Code in terms of critical features like seeing what's going on in subagents. The parent + subagent workflow has become critical for keeping the reasoning agent (parent) context-lean while also lett", "context": "", "use_case": "planning and reasoning", "switched_from": "", "story": "GPT-6 Sol and Luna"}, {"id": 796887649, "item_id": 49805478, "url": "https://news.ycombinator.com/item?id=49805478", "by": "yipinwong", "date": "2026-09-22", "stance": "neutral", "quote": "Also adding verification for that Fable 5.1 output in the same sesssion.", "excerpt": "I am sorry, I meant to say I generated STAR out of my resume line, trying to generate STAR, and polish it thus $5. --- I provided crapton of context for that one resume line. All the work I did, documentations for my justifications, etc. I initially messed up and came out ot $5, rest of resume used around $4 per line (I used a fresh new session on purpose). --- As a clarification, $2.2 average for OPUS 5.5 was the sa", "context": "", "use_case": "verification", "switched_from": "", "story": "Claude Opus 5.5"}, {"id": 796868720, "item_id": 49804295, "url": "https://news.ycombinator.com/item?id=49804295", "by": "port3000", "date": "2026-09-22", "stance": "recommend", "quote": "I use Fable as the orchestrator, only speak with Fable", "excerpt": "I believe Opus 5 isn't meant to be spoken to by humans. It's great at executing but I reckon it's intended to be spoken to by other models such as Fable. I use Fable as the orchestrator, only speak with Fable, and all implementation, recon, design etc happens with Opus 5, with Fable reviewing (and translating).", "context": "", "use_case": "orchestrating coding agents", "switched_from": "", "story": "Claude Opus 5.5"}, {"id": 796864993, "item_id": 49804062, "url": "https://news.ycombinator.com/item?id=49804062", "by": "visarga", "date": "2026-09-22", "stance": "warn", "quote": "The poor writing style made even models like Fable look worse than Sonnet 3.7 to me.", "excerpt": "I tried a lot to get rid of Claude vibes and I failed. I made blacklists, I used prompt evolution to test against good human writing samples, even created a dynamic policy for dialogue, a cli tool used to steer agents dynamically by well chosen questions. The poor writing style made even models like Fable look worse than Sonnet 3.7 to me.", "context": "", "use_case": "writing", "switched_from": "", "story": "I don't want to read what you didn't write"}, {"id": 796647731, "item_id": 49790483, "url": "https://news.ycombinator.com/item?id=49790483", "by": "swalsh", "date": "2026-09-21", "stance": "warn", "quote": "Fable chewed through my usage too quickly to be practical", "excerpt": "Codex has become my goto tooling. I used to be a Claude Max subscriber, but I was becoming disappointed with the quality of the output from Opus 5. Fable chewed through my usage too quickly to be practical. Moving to a Pro account w/ Codex was a big improvement. Sol had great output, and the usage was more than sufficient for most of my needs. However astra does tend to chew up usage, so when i've done to much of tha", "context": "", "use_case": "", "switched_from": "", "story": "Grok 4.7"}, {"id": 794875328, "item_id": 49679708, "url": "https://news.ycombinator.com/item?id=49679708", "by": "pdantix", "date": "2026-09-13", "stance": "recommend", "quote": "having fable plan and orchestrate with opus implemention + fable reviews, is my go-to.", "excerpt": "personally, i would not rely on opus 5 end to end as it'll start getting into walls of comment slop and shitting up the codebase similar to gpt 5.5's isRecord meme. on the other hand, having fable plan and orchestrate with opus implemention + fable reviews, is my go-to. if you give fable your guidelines up front or in your {claude,agents}.md, it will keep opus on a tight leash. opus can still write great code almost ", "context": "", "use_case": "planning and reviewing code", "switched_from": "", "story": "Real-SWE: Benchmarking AI models on private, real-world, enterprise codebases"}, {"id": 794870257, "item_id": 49679391, "url": "https://news.ycombinator.com/item?id=49679391", "by": "qskousen", "date": "2026-09-13", "stance": "recommend", "quote": "Fable 5.1 was noticeably better at getting things right and doing things in a good reliable way.", "excerpt": "I have a pet project I have been working away on for some time that involves building GPU backends for various cards in Zig, lots of complex stuff in it. Lately I mostly use Opus 5, it can pretty reliably plug away at things but it does mess stuff up occasionally. For this codebase, Fable 5.1 was noticeably better at getting things right and doing things in a good reliable way. Of course, I can only use Fable for a b", "context": "", "use_case": "tough coding tasks", "switched_from": "", "story": "Everyone should slow down AI development except for me"}, {"id": 794842177, "item_id": 49677636, "url": "https://news.ycombinator.com/item?id=49677636", "by": "ricardobeat", "date": "2026-09-12", "stance": "warn", "quote": "I don\u2019t have a use for Fable at current prices.", "excerpt": "Opus 5 at low effort is quite good at most tasks, and will go a long way even on the base plan. At higher thinking levels it is unbearably cautious and verbose - it drives me mad. I don\u2019t have a use for Fable at current prices. My current go-tos are DS Flash, Minimax M3 for subagents. Adding Muse Spark 1.3 to that mix but not convinced yet.", "context": "", "use_case": "", "switched_from": "", "story": "Ask HN: What default model do you use and why?"}, {"id": 794786592, "item_id": 49674162, "url": "https://news.ycombinator.com/item?id=49674162", "by": "jamesponddotco", "date": "2026-09-12", "stance": "recommend", "quote": "Fable is the only one that fucking listens to me and writes proper Go code", "excerpt": "Given that Fable is the only one that fucking listens to me and writes proper Go code, that\u2019s what I use for everything code-related. At least on my personal projects. If I don\u2019t care about the code, i.e. I\u2019m writing something quick just to test something, I\u2019ll use GPT Astra to save my Fable tokens. For conversations and research I usually go with GPT Astra Pro. For Home Assistant I use a combination of Grok 4.20 and", "context": "", "use_case": "code", "switched_from": "", "story": "Ask HN: What default model do you use and why?"}, {"id": 794771361, "item_id": 49673210, "url": "https://news.ycombinator.com/item?id=49673210", "by": "mariocesar", "date": "2026-09-12", "stance": "recommend", "quote": "I now use Fable to create a PLAN.md.", "excerpt": "I have an 'ask' alias in my shell that just uses Haiku. I use it daily for pretty much everything For more long work, I now use Fable to create a PLAN.md. I tell it to make a plan that will be executed by other models, and most of the time it ends up choosing Opus or Sonnet. I didn't start doing this recently. Before that, I would just use the top model for everything. Splitting the work across different models depen", "context": "", "use_case": "planning long work", "switched_from": "", "story": "Ask HN: What default model do you use and why?"}]}