Anthropic Launches Most Powerful Claude Models Yet at 20% Cheaper Price

Anthropic has introduced Claude Fable 5.2 and Claude Mythos 5.2, its most advanced AI models yet, claiming they outperform OpenAI's GPT-6 Astra in a number of benchmarks while costing 20% less.

Key points

• Claude Fable 5.2 scored 49.1% on FrontierCode 2.0, compared with 45.5% for Claude Opus 5.5 and 42.8% for GPT-6 Astra.
• On Humanity's Last Exam with tools, Fable 5.2 achieved 74.2%, ahead of Opus 5.5 at 67.7% and GPT-6 Astra at 57.2%.
• In its CVE Rediscovery evaluation, Mythos 5.2 detected 10.2% more vulnerabilities than Claude Opus 5.5 and 14.4% more than Mythos 5.1.
• Claude Fable 5.2 is available through Anthropic's platforms, including Amazon Web Services, Google Cloud, and Microsoft Azure.
• Confirmed on November 9, the new models cost $8 per million input tokens and $40 per million output tokens, down from $10 and $50 for their predecessors.

Confirmed on November 9, the new models cost $8 per million input tokens and $40 per million output tokens, down from $10 and $50 for their predecessors. This additionally makes them 20% cheaper than GPT-6 Astra at standard API rates. Cached input tokens cost $0.20 per million, compared with $1 for GPT-6 Astra, making them 80% cheaper.

For context, the lower rates and stronger noted performance in coding, reasoning, scientific research, and image understanding put Anthropic in a more competitive position against OpenAI. Nevertheless, Anthropic has not explicitly linked the rate reduction to competition with GPT-6 Astra.

Anthropic additionally claims the new models can generate responses up to 30% faster than their predecessors.

Both models employ the same underlying technology, but Fable 5.2 includes further cybersecurity and biological safety restrictions, allowing wider availability. Mythos 5.2 is aimed at approved organizations working in cybersecurity and life sciences. Anthropic Building Tool to Expose Chinese Models Secretly Using Claude.

Anthropic claims its new models outperform Claude Opus 5.5 throughout most tested capabilities and beat OpenAI's GPT-6 Astra in a number of benchmarks.

On Humanity's Last Exam with tools, Fable 5.2 achieved 74.2%, ahead of Opus 5.5 at 67.7% and GPT-6 Astra at 57.2%. The after table indicates selected benchmark results published by Anthropic. Benchmark Claude Fable 5.2 GPT-6 Astra Claude Opus 5.5. FrontierCode 2.0 49.1% 42.8% 45.5%. FrontierSWE v2 69.9% 65.5% 62.3%. AutomationBench 51.4% 41.4% 42.5%. Humanity's Last Exam 74.2% 57.2% 67.7%. OSWorld 2.1 89.3% 79.5% 87.2%. Chartography 73.4% 71.0% 64.4%.

Bold figures indicate the highest score in each benchmark. Results are based on Anthropic's published evaluations, generally using the highest available reasoning effort. Coding and Scientific Research.

Anthropic notes Fable 5.2 can handle complex software development tasks, including fixing bugs, migrating systems, and coordinating multiple AI agents.

During testing, an earlier version of the model supported migrate a production service from Python to Rust, making its startup approximately 26 times faster.

In another experiment, Fable 5.2 coordinated more than 1, 000 AI agents over eight days to develop a working operating system from scratch.

In practice, the firm additionally documented improvements in scientific research. Fable 5.2 supported solve two previously open theoretical physics difficulties involving black holes and completed a computer-verifiable formalization of Fermat's Last Theorem in 30 hours.

Meanwhile, Mythos 5.2 indicated improvements in protein structure prediction. Fable 5.2 additionally designed three working electronic devices and performed better in robotics evaluations. Claude Mythos 5.2 Focuses on Cybersecurity.

Anthropic describes Mythos 5.2 as its most capable model for cybersecurity engineering.

Notably, the firm claims the model can identify, reproduce, and assist fix more software vulnerabilities than previous Claude models.

While eligible life sciences researchers can apply through its Life Sciences Verification Program, cybersecurity professionals can request access through Anthropic's Cyber Verification Program.

Claude Enterprise customers can additionally access Mythos 5.2 capabilities through Claude Security. Anthropic Building Tool to Expose Chinese Models Secretly Using Claude.

Claude Fable 5.2 is available through Anthropic's platforms, including Amazon Web Services, Google Cloud, and Microsoft Azure. Anthropic has confirmed the after API pricing for both models. Pricing Category Standard Mode Fast Mode. Input tokens (1 million) $8 $16. Output tokens (1 million) $40 $80.

Speed improvement Up to 30% over 5.1 Up to 2.5x token generation.

For context, the optional fast mode is available through Claude Code, the Claude Platform, Google Cloud, and Microsoft Foundry. Nevertheless, access to Mythos 5.2's fast mode is limited to approved organizations.

Anthropic is additionally offering eligible paid subscribers further Fable usage during launch week. On Pro and standard Team seats, Fable can account for up to 50% of weekly usage limits.

Anthropic notes it tested the new models for cybersecurity threats, biological misuse, and attempts to escape restricted environments before release.

According to the firm, Fable 5.2 maintains robust safety protections while reducing unnecessary refusals compared with Fable 5.1.

Anthropic has additionally introduced further safeguards against attempts to copy its models through distillation.

While positioning Fable 5.2 and Mythos 5.2 for more demanding technical and research work, despite launching the new models, Anthropic continues to recommend Claude Opus 5.5 for most everyday coding and knowledge tasks. Stay Connected with ProPakistani.

Obtain the latest tech news, telecom insights, and product launches wherever you prefer. Follow on Google Discover. Follow on Google News Join WhatsApp. See more ProPakistani stories in Google Search and Top Stories.

Technology and Automotive Specialist covering the latest cars, smartphones, AI breakthroughs, and.

In short, anthropic Launches Most Powerful Claude Models Yet at 20% Cheaper Price is the central thread here, and readers can expect follow-up reporting as the picture becomes clearer.