#What Are the New AI Models from Anthropic and What Do They Offer?
Anthropic has just revealed two innovative AI models designed to maintain focus over extended tasks. Named Claude Fable 5 and Claude Mythos 5, these models launched on June 9 and target developers and advanced users. They are specifically built to excel in complex, multi-day projects, setting them apart from their predecessors.
#How Do the Pricing and Costs Compare?
The pricing of these models is noteworthy. At $10 for every million input tokens and $50 for output tokens, their rates are significantly lower than the previous Mythos Preview. This pricing strategy positions Anthropic uniquely in a market where other artificial intelligence companies typically increase costs for upgraded models. By lowering prices, Anthropic challenges the norm while promoting extended task performance.
#What Enhancements Are Introduced with Fable 5?
Feedback from developers highlights Fable 5's exceptional capabilities in long-horizon reasoning and intricate coding tasks. It has surpassed the 90% success rate on Anthropic’s core benchmark for complex operations, a figure its predecessor, Opus 4.8, found challenging to maintain consistently during extended use. With Fable 5, users can expect dedicated performance without needing to worry about session fatigue.
The model is generally available and includes safety features that ensure responsible usage. Users can access it via the Claude API, Claude.ai, and major cloud platforms including AWS Bedrock, Google Cloud, and Microsoft Foundry, making it easily integrable into existing systems without significant upheaval.
#What About Mythos 5?
Mythos 5 is available to select partners without the same safety measures as Fable 5, marking a strategic approach by Anthropic. This allows the company to offer a more flexible solution for partners while maintaining a commitment to safety with its accessible version. This dual offering puts Anthropic in a competitive position without compromising on their safety-first branding.
#Why Is the Competitive Landscape Significant?
The decision to price Fable 5 lower aims to attract power users who require comprehensive, long-duration output from AI services. The previous Mythos Preview was priced at over double the current rates. Developers, once limited by API call expenses, can now expand their use, allowing for more algorithmic explorations without constantly monitoring costs.
Given the widespread availability across major cloud platforms, teams on Azure, AWS, or Google Cloud can easily test Fable 5, facilitating innovation without incurring high reconfiguration costs. Overall, these developments begin to reshape how AI can be harnessed for complex tasks, offering a robust solution within an increasingly competitive market.