# fable Published articles for fable. This is one page of public article previews, not the complete archive. Follow Next page to continue. Summaries are not the original full articles. ## AI's best coding agent fails 60% of the time -- and the data backs it up DevFeed: [AI's best coding agent fails 60% of the time -- and the data backs it up]() Original publisher: [Read original article]() Author: Amanda Caswell Published: 2026-09-14T22:22:27Z Content type: news Language: en Sources: [The New Stack]() Topics: [Benchmark](), [AI-assisted coding](), [Fable](), [Claude Code](), [codex](), [Command-line interface](), [anthropic]() Tags: [ai](), [ai-engineering](), [ai-models](), [benchmark](), [benchmarks](), [claude-code](), [cli](), [codex](), [coding](), [fable](), [performance](), [software-testing]() ### AI overview Real-SWE evaluates coding agents on private company codebases and reports substantially lower success rates than public-repository benchmarks. Claude Fable 5.1, running through Claude Code, led the comparison with a 38.8% score, while the tested systems often failed most attempts. ### Source excerpt Claude Fable 5.1 just won a new coding benchmark despite failing more than six out of 10 times. Its 38.8% The post AI's best coding agent fails 60% of the time -- and the data backs it up appeared first on The New Stack. ## How Fable and Astra Could Change Software Engineering Work DevFeed: [How Fable and Astra Could Change Software Engineering Work]() Original publisher: [Read original article]() Author: Hemant Pandey Published: 2026-09-09T14:09:16Z Content type: opinion Language: en Sources: [The Hustling Engineer]() Topics: [Artificial Intelligence](), [coding](), [Fable](), [computer-use](), [Code](), [anthropic](), [OpenAI]() Tags: [ai](), [anthropic](), [code](), [coding](), [computer-use](), [engineering](), [fable](), [openai](), [terminal](), [vs-code]() ### AI overview This opinion article examines what Fable and Astra could change for software engineers. It contrasts Fable's long-running, multi-step coding and research work with Astra's ability to operate software and complete tasks across tools used by engineers. ### Source excerpt What actually changes for software engineers after Fable and Astra? ## Testing Astra 6 v Fable 5.1 on a Gradle docs bug DevFeed: [Testing Astra 6 v Fable 5.1 on a Gradle docs bug]() Original publisher: [Read original article]() Author: Laura Kassovic Published: 2026-09-09T04:00:00Z Content type: article Language: en Sources: [The Gradle Blog]() Topics: [Gradle](), [Testing](), [ai-coding](), [Benchmark](), [Fable](), [Documentation](), [bug]() Tags: [ai-coding](), [benchmark](), [bug](), [docs](), [fable](), [gradle](), [testing]() ### AI overview An engineering blog compares Astra 6 and Fable 5.1 after each agent fixes the same Gradle Kotlin DSL documentation bug. Both fixes passed independent judging, while the comparison examines cost, speed, and maintainability. The author emphasizes that the result is based on one run per model, one judge, and one bug. ### Source excerpt On the afternoon of September 7th, claude-fable-5-1 was handed a git repository, a GitHub issue, and fifty turns to fix it. Just over three hours later, gpt-6-astra finished the same assignment in another worktree. Same issue. Same constraints. Different CLI wrapped around each model: claude-code for Fable, codex for Astra. Both Agents fixed the bug. Both results were graded as successes by an independent judge. But these results were quite different. You would be forgiven for wondering why an engineering blog is running a two-model bakeoff on a single documentation bug, in a year when everybody and their intern has published an AI coding benchmark. Here's the honest answer: we wanted to know which one actually did the better job, as a side quest of the Agentic Gradle project. But "better" was never going to mean pass-or-fail. Both of these agents passed; if that were the whole story, this post would be four sentences long. What actually separates a fix worth merging from a fix worth sending back is cost, speed, and whether the result is something we would want to maintain, and those three do not all point the same direction, as you are about to see. So: one bug, two agents, and a very literal stopwatch. Here is what this looks like when you actually read the agents' transcripts. Real talk up front: this is n=1. One run per model, one judge, one bug. That's a thin base for anything you'd call a leaderboard, and I'll say so again at the end, but it didn't stop me from landing on an opinion by the time I'd finished reading both transcripts. The bug, briefly Issue #34751, filed August 21st by our very own cobexer, is the kind of bug that is very easy to describe and mildly annoying to fix. In the Kotlin DSL API docs, a type like Attribute linked to the Java 21 Javadoc. It should have linked to Java 17, because Gradle targets Java 17. The reason is almost embarrassingly mundane once you see it. Gradle is built with a JDK 21 toolchain but targets JDK 17, which i ## GPT-6 Astra and Fable 5.1: Starting and Reviewing Self-Directed Work DevFeed: [GPT-6 Astra and Fable 5.1: Starting and Reviewing Self-Directed Work]() Original publisher: [Read original article]() Author: Nate Published: 2026-09-06T15:03:01Z Content type: opinion Language: en Sources: [Nate's Substack]() Topics: [gpt-6-astra](), [Fable]() Tags: [fable](), [gpt](), [gpt-6-astra]() ### AI overview The article discusses changes involving GPT-6 Astra and Fable 5.1, focusing on starting self-directed work with minimal input, how long it can run without supervision, and the role of the user when it returns. ### Source excerpt What changed this week, with GPT-6 Astra and Fable 5.1, is how little you have to say to start one, how long it runs without you, and what your job is when it comes back. ## Claude Fable 5.1: Benchmark results and reasoning-effort experiments DevFeed: [Claude Fable 5.1: Benchmark results and reasoning-effort experiments]() Original publisher: [Read original article]() Author: Simon Willison Published: 2026-09-01T23:57:28Z Content type: article Language: en Sources: [Simon Willison]() Topics: [Fable](), [Benchmark](), [coding](), [Large Language Model](), [SVG]() Tags: [ai](), [ai-2-235](), [anthropic](), [anthropic-336](), [benchmark](), [benchmarks](), [claude](), [claude-310](), [coding](), [fable](), [generative-ai](), [generative-ai-1-981](), [llm](), [llm-reasoning](), [llm-reasoning-103](), [llm-release](), [llm-release-231](), [llms](), [llms-1-947](), [models](), [pelican-riding-a-bicycle](), [pelican-riding-a-bicycle-142](), [reasoning](), [svg]() ### AI overview The article examines Claude Fable 5.1 through benchmark results and a pelican SVG-generation experiment across its five reasoning-effort levels. It reports that low and medium settings appeared to skip reasoning for this prompt, while higher effort used more tokens, time, and cost. ### Source excerpt Today is Claude Fable (and Mythos) 5.1 day. Anthropic say that Fable 5.1 "sets a new standard for coding, knowledge work, and long-running problem-solving tasks". Their announcement spends a notable amount of time on scientific research, boasting of a 52.6% score on the brand new Terminal-Bench-Science 0.1 benchmark (first announced on August 27th), up from 24.7% for Fable 5, 29.0% for Opus 5 and 22.4% for GPT-5.6 Sol. Other benchmarks show slightly improved scores, but none as impressive as the Science one. But how well can it pelican? Back in July I wrote about how I was losing faith in the pelican benchmark - its connection to how good the models were at other tasks didn't seem to hold as strongly as it did back in 2025. The most interesting insights I get from it now are comparisons within model families, and particularly comparisons for the same prompt at different reasoning effort levels. Fable 5.1 has five reasoning levels: low, medium, high, xhigh, max - and no option to turn off reasoning entirely. I fixed an issue in llm-anthropic which caused reasoning traces not to be correctly recorded, then ran some prompts. Here's the full set of pelicans for all of the reasoning levels, each with the full reasoning transcript. I'll replicate them here: Low and medium, both without reasoning? Next, a bit of a mystery. This is what I got for effort low: The transcript doesn't show any summarized reasoning tokens, and the output token count is 1,998. With Claude that output token count includes reasoning tokens. It took 23.8 seconds and cost 10.017 cents. I bumped that up to medium and got this: Weirdly, that one also shows no reasoning text and used 1,977 output tokens - 21 tokens less than low. It took 23 seconds and cost 9.912 cents. So for this particular prompt ("Generate an SVG of a pelican riding a bicycle") Fable 5.1 appeared to skip reasoning entirely at both low and medium settings. High Here's high - 29.6 seconds, 2,612 output tokens, 13.087 cents: This one d ## One-shotting a Raccoon Heist game using Claude Fable 5 DevFeed: [One-shotting a Raccoon Heist game using Claude Fable 5]() Original publisher: [Read original article]() Author: Simon Willison Published: 2026-08-05T19:42:38Z Content type: tutorial Language: en Sources: [Simon Willison]() Topics: [Claude](), [Claude Code](), [GitHub Pages](), [GitHub](), [Code]() Tags: [ai](), [ai-2-235](), [anthropic](), [anthropic-336](), [browser](), [claude](), [claude-310](), [claude-code](), [claude-mythos-fable](), [claude-mythos-fable-41](), [code](), [coding-agents](), [coding-agents-248](), [fable](), [game-design](), [game-design-22](), [generative-ai](), [generative-ai-1-981](), [github](), [github-pages](), [iphone](), [llms](), [llms-1-947](), [prompt-engineering](), [prompt-engineering-197](), [text-to-image](), [text-to-image-47](), [vibe-coding](), [vibe-coding-97]() ### AI overview The article describes an experiment in which Claude Fable 5, running through Claude Code for web, built a Raccoon Heist game from an older GPT-3 game concept and screenshots. It also explains using GitHub Pages to preview work from Claude Code for web and notes that published content from a private repository may be viewable by anyone who guesses the repository name. ### Source excerpt Back in 2022 I tweeted screenshots of a game concept generated by GPT-3 and some concept "art" created using DALL-E. Today, on the fourth anniversary of that tweet, I decided to see if Claude Fable 5 (running in Claude Code for web) could build the entire game from the content of that tweet. It did a pretty good job of it! You can play the game here. Here's the GitHub repo, and a short video demo: Your browser does not support HTML5 video. How I built this This is the August 5th, 2022 tweet: My GPT-3 prompt back then was: Write a detailed product description of a computer game where a team of raccoons go on heists. This was in the text completion days, where you asked GPT-3 to complete the rest of the text. It completed: In "Raccoon Heist", you and your team of thieving raccoons are tasked with pulling off a series of daring heists. From robbing banks to stealing priceless art, no job is too big or too small for your furry crew. You'll need to use your wits and your skills to avoid the police and make a clean getaway with the loot. With exciting gameplay and a charming cast of characters, "Raccoon Heist" is the perfect game for anyone looking for a light-hearted caper. My prompt to DALL-E was short: Screenshot from a video game where a team of raccoons go on a heist Today's experiment: can I dump those screenshots into Fable 5 with a prompt telling it to write a game, then leave it to its own devices and get a working game at the end? Setting Claude Code for web up to use GitHub Pages A frustrating thing about Claude Code for web is that it can be hard to test what it's working on while it's still working. I've been using GitHub Pages to work around that limitation, and found it to work really well. Here's my process: Create a new repository for the project at https://github.com/new - this can be public or private, the trick works equally well for both. Start a Claude Code for web session, in the Claude iPhone or Desktop apps or in the browser at https://claude.ai/c ## Cost-effective orchestration with Fable 5 DevFeed: [Cost-effective orchestration with Fable 5]() Original publisher: [Read original article]() Author: Jeff Morhous Published: 2026-07-21T11:37:27Z Content type: tutorial Language: en Sources: [The AI-Augmented Engineer]() Topics: [Fable](), [Claude Code](), [Orchestration]() Tags: [better](), [claude-code](), [cost](), [effective](), [fable](), [how-to](), [results]() ### AI overview This tutorial presents Fable 5 in Claude Code as an orchestration approach and claims it can produce better results at lower cost. The supplied text does not provide implementation details or supporting measurements. ### Source excerpt Fable 5 in Claude Code is effective. I'll show you how to use Fable as an orchestration to get better results at a fraction of the cost ## Bun's migration from Zig to Rust reduced the estimated timeline from 1-2 years to 11 days DevFeed: [Bun's migration from Zig to Rust reduced the estimated timeline from 1-2 years to 11 days]() Original publisher: [Read original article]() Author: Ivan Klaric Published: 2026-07-16T16:50:20Z Content type: opinion Language: en Sources: [The Pragmatic Engineer]() Topics: [Bun](), [migration](), [Rust](), [Zig](), [Fable](), [Memory Leaks]() Tags: [bun](), [fable](), [memory-leaks](), [migration](), [rust](), [zig]() ### AI overview The article examines Bun's rewrite from Zig to Rust, focusing on memory-safety problems and the migration approach. It reports that the migration timeline was shortened from an estimated 1-2 years to 11 days, while emphasizing that thorough testing was required. ### Source excerpt To a sceptic, spending $165K to migrate Bun from Zig to Rust sounds very expensive. But to a realist, shortening a 1-2 year migration down to 11 days opens amazing new opportunities for devs. However, a thoroughly-tested project is required to pull it off. ## Athena's free vulnerability intelligence model and the economics of security clearinghouses DevFeed: [Athena's free vulnerability intelligence model and the economics of security clearinghouses]() Original publisher: [Read original article]() Published: 2026-07-15T00:00:00Z Content type: opinion Language: en Sources: [Chainguard: Unchained]() Topics: [Security](), [Open Source](), [Critical Infrastructure](), [Network]() Tags: [akrites](), [chainguard-athena](), [critical-infrastructure](), [fable](), [infrastructure](), [money](), [mythos](), [network](), [open-source](), [security](), [software-supply-chain-security](), [trust](), [vulnerability-clearninghouse]() ### AI overview The article explains why Athena provides pre-disclosure vulnerability findings to infrastructure and network partners for free. It argues that allowing partners to commercialize services built on advance security information can align incentives, trust, and broader defensive value, while distinguishing government programs from commercial clearinghouses. ### Source excerpt Why give away the most valuable data in security? Learn how Athena's business model aligns trust, incentives, and open source defense. ## Which Doc Format is Best for AI Specifications? DevFeed: [Which Doc Format is Best for AI Specifications?]() Original publisher: [Read original article]() Author: Peter Lawrey (noreply@blogger.com) Published: 2026-07-14T14:00:28Z Content type: article Language: en Sources: [Vanilla Java]() Topics: [Specifications](), [Markdown](), [Fable](), [HTML](), [Documentation]() Tags: [ai](), [fable](), [html](), [info](), [markdown](), [readability](), [specifications]() ### AI overview The article compares Markdown, AsciiDoc, and HTML for AI specifications using document-size, markup-overhead, readability, and editing-workflow observations. It recommends Markdown for AI working documents, AsciiDoc for curated human-reviewed specifications, and HTML primarily as a publishing target. ### Source excerpt I have a Spec Driven Project with 674 documents incl 181 specs, 429 tasks, and 40 project docs. About 1/3 of these are an acceptance sub-project to show the root project meets fit-for-purpose requirements. I used Fable to convert these to and from AsciiDoc, Markdown, and HTML. From this, I tried to draw some conclusions about which format is better than the others for AI specifications. TL;DR: Markdown for AI working documents, AsciiDoc for curated human-reviewed specs, HTML only as a publishing target. Objective Metrics Corpus Tokens (Approx) Markdown: 594k, AsciiDoc: 598K, HTML: 661k Worst Single-File Ratio AsciiDoc: 1.023x, HTML: 1.26x (due to tables and code blocks) Markup share of file content Markdown 0.9%, AsciiDoc 1.3%, HTML 9.9% - a 10x overhead for HTML Read/write Assessment In terms of readability, both Markdown and AsciiDoc are good. HTML has far more tags, which add noise. One of the most common tags is some code however in the other formats, this is just `some code` When you have an average of 16 per file, this adds a lot of noise. Usability GitHub renders both Markdown and AsciiDoc with working links and checkboxes; HTML is shown as raw source, so the review loop dies there. The first two are also easier to view while you're editing in IntelliJ. HTML is more difficult to read and edit, and it is not as easy to view the rendered output. If you occasionally need HTML, you don't need HTML files: embed an island with ```{=html} in Markdown (Pandoc raw-attribute syntax) or in AsciiDoc. It renders where HTML is supported and drops out cleanly elsewhere. If you really need to use HTML, you can include a link to a file in that format. Note As these were one-for-one conversions, I didn't use features that only exist in AsciiDoc or HTML, such as. Markdown vs AsciiDoc While you might choose either format for your specs, I have been following this convention to keep the authoring consistent. Markdown used for AI-generated content, not human-written, ## Claude Fable 5 access restored on AI Gateway DevFeed: [Claude Fable 5 access restored on AI Gateway]() Original publisher: [Read original article]() Author: Jerilyn Zheng Published: 2026-07-01T07:01:00Z Content type: release Language: en Sources: [Vercel News]() Topics: [Fable](), [Anthropic Claude](), [API](), [Artificial Intelligence](), [debugging]() Tags: [anthropic-claude](), [api](), [coding](), [debugging](), [docs](), [fable](), [government](), [model](), [models](), [real-time](), [retention](), [safety]() ### AI overview Access to the Claude Fable 5 Mythos-class model has been restored on AI Gateway after the US Government lifted export controls. The release updates its safety classifiers and documents model fallbacks, API usage, and 30-day retention of prompts and completions. ### Source excerpt Access to Claude Fable 5, the Mythos-class model, has now been restored on AI Gateway following the US Government's decision to lift the export controls. Fable 5 is the same model that was available between June 9 and June 12. What has changed is the safety classifiers, which are now updated and more robust. In the near term, some routine tasks such as coding and debugging may trigger safety classifiers. To ensure requests are still serviced when the safety classifiers are triggered, use model fallbacks. AI Gateway will try each model in models in the stated order if Anthropic refuses the request to Fable 5. To call Fable 5, use model name anthropic/claude-fable-5: Model fallbacks work on every API format: for more information on how to configure these, see the docs. Anthropic does not support Zero Data Retention for the model, because some misuse patterns are only visible across cumulative requests, which real-time filters cannot catch on their own. Prompts and completions are retained for 30 days and are not used to train Claude. Read more in the data retention whitepaper. Read more ## Scaling Context Management: Comparing Fable 5 and GPT-5.5 DevFeed: [Scaling Context Management: Comparing Fable 5 and GPT-5.5]() Original publisher: [Read original article]() Author: Calvin French-Owen Published: 2026-06-13T12:00:00Z Content type: opinion Language: en Sources: [Calvin French-Owen]() Topics: [long-context](), [codex](), [Fable](), [API](), [Caching](), [GPU]() Tags: [api](), [caching](), [codex](), [context](), [fable](), [gpu](), [long-context]() ### AI overview The article compares how Fable 5 and GPT-5.5 scale context management for difficult, long-running tasks. It describes Codex's server-side compaction approach and discusses its implications for clients and long-running threads. ### Source excerpt Comparing two very different approaches to scaling context between Fable 5 and GPT 5.5.