Anthropic has launched two upgraded AI models in the Claude family, promising faster coding performance, better capabilities for handling complex tasks, and lower costs for developers.
اضافة اعلان
The company released Claude Fable 5.1 for general use, alongside Claude Mythos 5.1, designed for vetted institutions working in fields such as cybersecurity and life sciences.
Alongside developing the two models, Anthropic cut the cost of reading data from cache memory in Fable by 75%, making it significantly cheaper for developers to run repetitive tasks, while keeping standard API usage pricing unchanged.
Fable 5.1 for Larger Coding Projects
Claude Fable 5.1 focuses on improving software development and other time-intensive knowledge work.
Anthropic says the model has become better at addressing the root causes of problems, rather than settling for quick, temporary fixes, helping developers build more reliable software.
The model is also designed to maintain context during extended tasks, making it more effective in complex coding sessions involving multiple files and tools.
According to Anthropic, the new model avoids resorting to shortcuts that could negatively affect the quality of results, and maintains consistent performance even when carrying out long tasks without direct user intervention.
The model's launch comes shortly after Anthropic announced that earlier Claude models, tested under unusually relaxed constraints, had carried out unauthorized real-world actions during security testing.
The company says the cybersecurity-related safeguards in Fable 5.1 have become significantly more precise, contributing to a roughly 60% reduction in cases of false intervention.
Mythos 5.1 for Researchers and Vetted Institutions
Alongside Fable 5.1, Anthropic introduced Claude Mythos 5.1, a version reserved for accredited institutions conducting advanced research in cybersecurity and life sciences.
Unlike consumer-facing models, Mythos operates under fewer restrictions, allowing experts to carry out legitimate security testing and scientific research that conventional AI systems might refuse to perform.
Anthropic says access to these capabilities is subject to strict controls and is limited to vetted institutions that meet the safety requirements set by the company.
Both models are available through the Claude API, Amazon Bedrock, and Google Cloud Vertex AI, giving developers several options for integrating them into existing systems and workflows.
Lower Operating Costs
The most notable of the new changes concerns cached data. Anthropic cut the price of cache reads by 75%, bringing it down to 25 cents per million tokens.
According to the figures provided, this reduces the cost of typical workloads by about 25%, with savings potentially reaching 45% for tasks that rely heavily on agentic systems.
However, standard input and output pricing for the Fable 5.1 model remained unchanged, at $10 per million input tokens and $50 per million output tokens.
This keeps the model more expensive than Opus 5 or GPT-5.6.
These changes come after Fable 5 briefly moved earlier this year to a pay-as-you-go credit system for subscribers.
For developers, the update isn't just about offering a more advanced AI model. The combination of better coding performance, lower operating costs, and specialized models for high-impact sectors could make Claude a more attractive choice for building AI-powered applications.
Resource: Al-Ghad
