What is Claude Fable 5

Claude Fable 5 is a powerful artificial intelligence model released by Anthropic on June 9, 2026. It belongs to a special group of models called Mythos-class, which are designed to handle very complex tasks. This version is meant for general use by anyone, while a more restricted version called Claude Mythos 5 is available only to trusted partners. The model is built to excel in software engineering, knowledge work, visual analysis, and scientific research. It represents a major step forward in AI performance compared to previous versions.

Benefits

Claude Fable 5 offers several key advantages that make it a strong tool for professionals and researchers. Its most notable strength is in software development. The model can handle long and complex coding projects with high reliability. It can even rebuild entire web applications just by looking at screenshots. In one test, it successfully played through a video game using only images from the game screen. This ability to work autonomously means users need to give fewer instructions to get results.

The model also shows great improvement in analytical reasoning. It can break down complex data and finish tasks much faster than earlier models. It is particularly strong in finance, handling both general financial tasks and complex reasoning with ease. For visual tasks, it can extract precise numbers from scientific charts and perform complex visual analysis. It can even write code to create fluid simulations that sync with music it has never heard before.

In the field of science, the restricted Mythos 5 version has shown transformative potential. It has accelerated the drug design process by ten times. In some studies, it matched or beat skilled human scientists in designing proteins. It can also generate new scientific hypotheses that experts find compelling. One of its hypotheses about a bacteria protein was later confirmed in a major science journal. The model is also very efficient at writing scientific papers, creating drafts that are much shorter than usual.

Safety is a major focus for this model. It includes built-in safeguards to prevent misuse in areas like cybersecurity, biology, and chemistry. If a safety filter is triggered, the system automatically switches to a capable backup model to ensure a helpful response. Extensive testing found no universal ways to bypass these safety measures. The model also follows a strict data policy where customer data is kept for only 30 days and is never used to train new models.

Use Cases

Claude Fable 5 is designed for a wide range of applications. Software engineers can use it to build complex applications with fewer steps. Developers can feed it screenshots of a website, and it can generate the full source code. Data analysts can use it to process large spreadsheets and financial reports quickly. The model finishes these tasks significantly faster than previous tools.

Researchers in life sciences can use the Mythos 5 version to design new drugs and proteins. It can run through the entire scientific workflow, from choosing binding sites to recovering from errors. Scientists can use it to generate new ideas for experiments that they might not have considered on their own. The model can also help write scientific papers by generating concise drafts that highlight key findings.

Visual analysts can use the model to interpret complex scientific figures and diagrams. It can pull out exact numbers and understand relationships between different visual elements. Creative professionals can use its ability to generate code based on visual input to create unique art or simulations. For example, it can write code to create visual effects that match a specific musical rhythm.

Pricing

Anthropic has set competitive pricing for these models. Both Claude Fable 5 and Claude Mythos 5 cost $10 per million input tokens and $50 per million output tokens. This price is less than half of what the previous Claude Mythos Preview cost. For businesses, there are Enterprise plans available that are based on how much the model is used. When the models first launched, they were included in subscription plans like Pro, Max, and Team at no extra cost. After a short period, access on these plans required a specific upgrade.

Vibes

Early feedback from customers has been very positive. Users have highlighted the model's ability to solve problems that were previously out of reach for earlier AI tools. The capacity to handle long-horizon projects without constant human intervention has been a major win. In the scientific community, the ability to generate novel hypotheses that experts prefer over those from other models has generated excitement. The fact that one of its hypotheses was published in a top journal adds to its credibility. The safety measures have also been well received, as they allow powerful capabilities without the risk of misuse.

Additional Information

The launch of these models was temporarily suspended on June 12, 2026, due to a US government export control directive. Anthropic is working to restore access as soon as possible. The company invested heavily in safety, including a 1,000-hour external bug bounty program to find any weaknesses in the system. This rigorous testing ensures that the models are safe for public use. The data retention policy, which limits how long customer data is stored, sets a new standard for privacy in high-capability AI models.