Tag
Superpowers 6 is released, using Fable 5 to run 25 autonomous experiments, improving build speed by 50% and reducing token costs by 60%, with detailed records of the experimental process and lessons from failures.
The same account first said Fable 5 had gotten weaker, then posted that Fable 5 is back, sharing a one-shot Minecraft-style demo showcasing its scene, lighting, and inventory capabilities, indicating the model's capability hasn't been fully removed.
This tweet introduces a Fable 5 Prompt structure that is suspected to be internally optimized by Anthropic, emphasizing that writing prompts should focus on goals, boundaries, and acceptance criteria rather than lengthy instructions. Especially for stronger models like Claude and Codex, avoid teaching the model how to think.
Discusses Fable 5's pricing at twice Opus, currently free via Claude subscription until July 7, after which pay-per-usage, with switching advice.
Someone used the Fable 5 model to generate a complete ink-wash style Roguelike game "Ink Guard" within five hours, including visuals, music, boss fights, and self-testing mechanisms, demonstrating the ability of AI to independently complete game development.
Anthropic confirms that access to Fable 5 via subscription ends on July 7, after which it will move to usage credits, but plans to restore it as a standard subscription model when capacity improves.
A user reports that Fable 5's new classifiers misrouted 75% of a coding session to Opus, flagging routine coding as a cybersecurity risk and causing unexpectedly high costs.
Fable 5's chain-of-thought reasoning was accidentally exposed in its web interface, revealing a rambling but endearing internal monologue.
Fable 5, a frontier model, returned with new safety guardrails that significantly degrade its performance on debugging, refactoring, and hallucination benchmarks, routing flagged requests to a less capable model (Opus 4.8), marking the end of an era for unrestricted access to advanced AI.
Tip to use Fable-5 as orchestrator and Codex as implementer to save credits, with a link to a library of composable AI agent skills for autonomous software development.
Fable 5 model generates a 644-line single-file HTML from a one-sentence description, achieving a realistic American flag fluttering effect including folds, lighting, and rhythm, all in one shot without any revision, demonstrating the AI's capability in complex code generation.
Fable 5 model only used 4 prompts and $173 worth of tokens to create a game called 'Super Smart Racing', demonstrating its extremely strong generative capabilities.
Sonnet 5 的基准测试不如 Opus 4.8，但通过 Claude Code 中的三步设置（/model Sonnet 5，/effort Ultracode，复杂任务自动触发 Dynamic Workflow）可以充分发挥其优势，配合 Fable 5 作为智能顾问，实现模型分工协作。
Fable 5 resumed after a 19-day suspension due to US government export controls, with some changes during that time.
Anthropic announced the return of Fable 5 with a pixel monster animation video, using a light and creative approach to conclude the 19-day export control turmoil.
Fable 5 is a model router for Opus models, noted as the most expensive ever constructed.
A tweet thread explaining how to configure Fable 5 as the orchestrator with Opus and Sonnet as subagents, plus Codex as a peer engineer, in Claude Code to optimize model usage and task delegation.
A user questions why widespread white-collar unemployment hasn't occurred despite frontier models like Fable 5 and GPT 5.6 benchmarks showing capability for most tasks, and seeks clarification on remaining bottlenecks and adoption realities.
Anthropic reinstated Claude Fable 5 after US export restrictions were lifted, and is collaborating with Amazon, Microsoft, Google on a shared framework to assess AI jailbreak severity across the industry.