Anthropic's new AI model, Claude 3.5 Sonnet, outperforms OpenAI's GPT-4o in various benchmarks

From Verizon Media: 2024-06-20 13:01:36

Anthropic launched Claude 3.5 Sonnet, a new AI language model that surpasses the previous model, Claude 3 Opus, while operating twice as fast. Sonnet is the first release in the Claude 3.5 family, with Haiku and Opus to follow later this year. The new model excels in understanding nuance, humor, and complex prompts.

Anthropic claims Sonnet sets industry records in graduate-level reasoning, undergraduate-level knowledge, and coding proficiency. It outperforms OpenAI’s GPT-4o on multiple benchmarks. The model also shows better visual input interpretation compared to Claude 3.0 Opus, making it valuable for sectors like retail, logistics, and financial services that require data interpretation from images.

The company introduced Artifacts, a new workspace for Claude users to generate content and collaborate with team members. It allows users to prompt the chatbot and make changes in a dedicated window. Anthropic envisions Claude becoming a hub for team collaboration by centralizing knowledge, documents, and ongoing work in a shared space.

Claude 3.5 Sonnet is now available for users to try on the Anthropic website, Claude iOS app, and through API integration with platforms like Amazon Bedrock and Google Cloud’s Vertex AI. Pricing remains the same as the previous model at $3 per million input tokens and $15 per million output tokens. Users on Pro and Team subscriptions receive higher token counts.



Read more at Verizon Media: Anthropic’s newest Claude chatbot beats OpenAI’s GPT-4o in some benchmarks