What is Claude Fable 5? 7 things to know about Anthropic's first Mythos model

After months of warning about the potential dangers of releasing Mythos, Anthropic is bringing its popular model to the public but with a few caveats. The new model is actually called Fable 5 and comes with restrictions in areas where Anthropic think it could be misused.

However,despite the restirctions Fable 5 is claimed to be the top of the line AI model available not just from Anthropic but from any AI provider at the time.

Here are top 7 features of Fable 5 should know:

1) Fable 5 is Anthropic's most powerful public model yet

Anthropic says Fable 5 outperforms any previous Claude model it made so far and has achieved a state-of-the-art performance on nearly all tested AI benchmarks including software engineering, vision, scientific reasoning and knowledge work.

2) Fable 5 takes a lead over GPT-5.5 and Gemini 3.1 Pro

Anthropic also says that Fable 5 outperforms rival AI models from OpenAI and Google across a variety of benchmarks.

On SWE-Bench Pro, a benchmark that measures how well AI models can solve real-world software engineering problems, Anthropic says Fable 5 scored 80.3% while both GPT-5.5 (58.6%) and Gemini 3.1 Pro (54.2%) ranked even behind Claude Opus 4.8 (69.2%).

Similarly, on FrontierCode (Diamond), a benchmark that measures real-world production quality of AI generated code, Fable 5 scored 29.3% while GPT-5.5 lagged behind at 5.7%.

Fable 5 also took the lead in the coveted Humanity's Last Exam benchmark which measures the limits of AI reasoning and depth of scientific knowledge. Fable 5 came out on top with a score of 59% without tools use and 64.5% with tools while GPT-5.5 and Gemini 3.1 Pro fell behind with a score of 41.4% and 44.4% respectively.

3) Fable 5 can compress months of coding into days

Anthropic says that during early testing with Stripe, Fable 5 was able to compress months of software engineering into days. The AI model is said to have completed a 50-million-line Ruby codebase migration in a single day that would otherwise have required a team of engineers working for more than two months.

4) Rebuild apps from screenshots

Anthropic says Fable 5 is also excellent for tasks that require vision capabilities. This means that the model can extract information from scientific charts, understand complex diagrams and even recreate a web application's source code using screenshots alone

5) Fable 5 vs Mythos 5:

While Fable 5 and Mythos 5 are both being released together, Anthropic is severely placing restrictions on many areas about which Fable 5 cannot talk. Some of these restrictions are being lifted on Mythos 5 especially around cybersecurity, biological research and advanced scientific reasoning which the company will only make available to cybersecurity professionals and researchers

6) Guardrails preventing misuse:

In order to prevent misuse of Fable 5, the model automatically routes conversations around cybersecurity, biology and AI model distillation to its Claude Opus 4.8 model instead.

Anthropic, however, claims that 95% of Fable sessions don't need to fallback on Opus 4.8 at all.

7) Anthropic's AI is helping with drug discovery:

Anthropic claims in its blog that Mythos 5 is helping accelerate parts of drug design process by around ten times. The company says it model is capable to completing all stages of drug design that are usually performed by humans such as hoosing binding sites, selecting and running protein design tools, and recovering from failures

How to use Fable 5?

Fable 5 is available to use via the Claude app and website along with API. In order to use the new AI model, you will need a Pro, Max, Team or a seat-based Enterprise plan.

However, Anthropic says Fable 5 will remain included in these subscription plans only until 22 June and. From 23rd June, these users will have to begin paying for credits to access the powerful AI model.