Anthropic released updated AI models Fable 5.2 and Mythos 5.2, offering faster performance and lower costs as the company prepares for a major stock market debut.
Key facts
- •Fable 5.2 costs $8 per million input tokens, a reduction from the previous $10 price point.
- •The new models operate up to 30 percent faster than previous versions on complex tasks.
- •Mythos 5.2 is restricted to vetted organizations due to fewer safety restrictions compared to Fable 5.2.
- •Anthropic's third-quarter revenue reached $17 billion, up from $1.4 billion in the same period last year.
- •The company expects to raise up to $100 billion in its IPO, which could make it the largest on record.
Anthropic announced the release of two new AI models, Claude Fable 5.2 and Claude Mythos 5.2, on Monday. The company states that the new versions show improvements in coding, reasoning, and vision, and operate up to 30 percent faster than their predecessors. Fable 5.2 is available to all users, while Mythos 5.2 is restricted to vetted organizations due to fewer safety restrictions.
By the numbers
Model Capabilities and Pricing
Anthropic reports that Fable 5.2 can handle complex software engineering problems and orchestrate agent teams for long-running projects. During prerelease testing, the models helped engineers complete software migrations, design electronic devices, and solve open problems in theoretical physics, including formalizing the proof of Fermat’s Last Theorem in 30 hours. The cost of Fable has been lowered from $10 to $8 per million input tokens, making it cheaper than its primary competitor, GPT-6 Astra.
Safety and Access Restrictions
Fable 5.2 includes safeguards that block most cybersecurity and biology tasks, allowing it to be offered to the general public. In contrast, Mythos 5.2 has fewer restrictions and is available only to vetted organizations working in cybersecurity or biomedical research. Anthropic maintains that Mythos 5.2 does not have the ability to act covertly against users, though its system card notes it shows stronger covert capabilities than prior models when reasoning is hidden from monitors.
IPO Context and Financials
The release comes as Anthropic prepares for an initial public offering expected the week of November 16, potentially raising up to $100 billion. The company’s revenue reached $17 billion in the third quarter, up from $1.4 billion a year earlier, though it posted an operating loss of $567 million. The IPO is expected to value the company at over $1 trillion, serving as a gauge for the sustainability of the current AI spending boom.
Advertisement
This article was independently rewritten by ManyPress editorial AI from reporting originally published by The Verge AI, The Hindu Technology, TechCrunch AI.
