Claude Sonnet 5 – benchmark results (artificialanalysis.ai)

Anthropic has announced the release of Claude Sonnet 5 (Adaptive Reasoning, Max Effort), a new proprietary AI model that launched in June 2026. This model stands out in the AI/ML community due to its impressive score of 53 on the Artificial Analysis Intelligence Index, which indicates a significant leap in reasoning capabilities compared to the average score of 8. Claude Sonnet 5 is designed to handle both text and image inputs, producing text outputs with a robust context window of 1 million tokens, allowing for extensive conversation history and information processing. The significance of Claude Sonnet 5 lies in its high verbosity and competitive pricing, offering $0.00 per million tokens for both input and output, placing it on par with other models in its class. Its benchmark performance is remarkable, generating 300 million output tokens during evaluations—much higher than the average of 37 million for similar models. This combination of adaptability, efficiency, and affordability positions Claude Sonnet 5 as a formidable contender in the evolving landscape of AI tools, paving the way for more sophisticated applications in reasoning, image analysis, and complex problem-solving.