Anthropic Extends Claude Fable 5 Access After OpenAI Sol Launch

Anthropic has once again expanded subscriber access to its Claude Fable 5 model, extending availability through July 19 at no additional cost. The announcement, made Sunday through X, subscriber emails, and an updated support page, marks the company’s second expansion of Fable access within a single week as competition in the AI industry intensifies following OpenAI’s latest model release.

The latest extension follows a promotional offer introduced on July 7, which allowed Claude subscribers to use Fable for up to 50% of their weekly subscription limits. Anthropic also said users who reach those expanded limits can continue using the model by purchasing usage credits, enabling subscribers to pay based on their actual usage.

The move comes three days after OpenAI introduced GPT-5.6, led by its flagship model Sol. OpenAI described the new system as delivering “state-of-the-art” performance in coding, scientific research, cybersecurity and knowledge work while claiming it surpasses previous and competing frontier AI models at a lower cost. The company also asserted that Sol and its accompanying models outperform Anthropic’s Fable on coding benchmarks and said Fable does not directly handle certain scientific research prompts because it falls back to an earlier model for most biology and chemistry-related requests, a limitation Anthropic has previously acknowledged.

Anthropic has also outlined plans to make Mythos, the more advanced model on which Fable is based, available to a broader audience in the future. For now, access remains limited to around 150 organizations across more than 15 countries. Released to a small group of partners last month for defensive cybersecurity work, Mythos was described by Anthropic as capable of finding and exploiting software vulnerabilities more effectively than any other AI model and nearly all human security experts.

The rivalry between Anthropic and OpenAI has become increasingly visible in recent months. Anthropic CEO Dario Amodei, who spent nearly five years at OpenAI before co-founding Anthropic in 2021, said last month that he left because he did not share the same vision as OpenAI CEO Sam Altman, stating, “At the end of the day, why argue with someone when you don’t have the same vision and you don’t trust them?” Earlier this year, Altman criticized Anthropic over a Super Bowl advertisement, accusing the company of being “so clearly dishonest” and “authoritarian,” while insisting OpenAI had no plans to introduce advertising in the manner suggested by its rival.

Anthropic originally introduced Fable in early June as a restricted version of its more powerful Mythos model, incorporating safeguards designed to prevent users from seeking guidance on cybersecurity as well as biology and chemistry topics related to bioweapons. Days after its debut, the U.S. Commerce Department required the company to withdraw the model after researchers at Amazon demonstrated a method for bypassing some of its safety protections. According to Anthropic, the researchers showed that Fable could demonstrate how a software vulnerability might be exploited. The company said it has since blocked that technique in more than 99% of cases. Anthropic has also remained at odds with the Trump administration following the Pentagon’s decision in March to designate the company as a supply chain risk after a dispute over unrestricted government access to its AI models.