AI unicorn Anthropic announced the launch of Claude Fable 5 on June 9 (local time), its first publicly available Mythos-class model and the company's most powerful public release to date. The model is now accessible to paying users, while the latest version of the frontier model Claude Mythos 5 continues limited access through Project Glasswing to select cybersecurity defenders and infrastructure providers. Anthropic positions Fable 5 as outperforming all previously released public models across nearly all AI performance benchmarks, with superior capabilities in software engineering, knowledge work, vision, and scientific research.
Anthropric states that "the longer and more complex the task, the stronger Fable 5's lead over other models becomes."
Anthropic Sets Unified Pricing at $10 Per Million Input Tokens
Fable 5 and Mythos 5 share identical pricing: $10 per million input tokens and $50 per million output tokens. Due to anticipated high and unpredictable demand for Fable 5, Anthropic will require users to spend credits to access the model starting June 23, until computing capacity returns to sufficient levels.
Fable 5 Completes 50-Million-Line Codebase Migration in One Day
According to a usage report from fintech company Stripe, Fable 5 completed migration of an entire Ruby codebase containing 50 million lines of code in one day. The same task would require a team more than two months to complete manually.
Anthropric emphasizes that Fable 5 is currently the most advanced model for vision tasks. Using vision alone, the model successfully completed the video game Pokémon FireRed, whereas earlier Claude bots required relatively complex auxiliary equipment to play the game.
Anthropic Implements Auto-Downgrade Safety Mechanism for High-Risk Queries
To ensure safety, Fable 5 employs an "automatic safety downgrade" mechanism that performs automatic safety checks on every user request. If the system detects requests involving three high-risk red-line areas—offensive cybersecurity techniques, sensitive biological and life sciences queries, and high-risk jailbreak attempts—it automatically switches the conversation back to the previous generation flagship model Claude Opus 4.8 for response.
Dianne Penn, Head of Research and Lab Project Management at Anthropic, stated: "For application scenarios that do not involve cybersecurity, we want to launch Fable as quickly and safely as possible. Therefore, we decided to release this portion of capabilities first while continuing to advance broader cybersecurity application development work."
In April, Anthropic revealed that the company had developed a powerful new AI model called Mythos. Due to potential safety concerns, Mythos was released only within a limited scope.
Anthropic Files Confidential S-1 Registration Statement on June 1
Anthropric and competitor OpenAI are both actively advancing toward public offerings. Anthropic's latest funding round valued the company at $965 billion post-money, while OpenAI's post-money valuation reached $852 billion. Both companies' IPO valuations are expected to reach $1 trillion.
On June 1, Anthropic published an article on its official website stating that the company had confidentially submitted a draft S-1 registration statement to the U.S. Securities and Exchange Commission (SEC), formally initiating the IPO preparation process. After the SEC completes its review, the company can choose to go public. According to Anthropic's disclosure, the company's annual recurring revenue (ARR) surpassed $47 billion in early May and is expected to reach $50 billion by the end of this month.
On June 8, OpenAI announced that it had recently submitted a confidential S-1 registration statement. OpenAI stated that it has not yet determined a specific listing timeline, noting that remaining a private company may facilitate certain matters and that the listing process may take some time. According to foreign media citing sources familiar with the matter, OpenAI also plans to advance a tender offer allowing employees to sell shares at the latest valuation.
FAQ
What did Anthropic announce on June 9?
Anthropric announced the launch of Claude Fable 5 on June 9 (local time), its first publicly available Mythos-class model. The model is now accessible to paying users at $10 per million input tokens and $50 per million output tokens.
How does Fable 5's safety mechanism work?
Fable 5 uses an automatic safety downgrade mechanism that checks every user request. If the system detects requests involving offensive cybersecurity techniques, sensitive biological queries, or high-risk jailbreak attempts, it automatically switches the conversation to Claude Opus 4.8 for response.
When did Anthropic file its IPO registration statement?
Anthropric confidentially submitted a draft S-1 registration statement to the SEC on June 1. The company's annual recurring revenue surpassed $47 billion in early May and is expected to reach $50 billion by the end of this month.
