Anthropic, a small AI company founded by two brothers who previously worked at OpenAI, has announced Claude 3.5 Sonnet, a new language model that surpasses GPT-4o and Google’s Gemini.

Claude 3.5 Sonnet is not the most powerful model Anthropic has. It is more like the middle sibling in their lineup. The company also offers a less powerful model named Haiku and a more advanced one called Opus. Despite this, Claude 3.5 Sonnet is capable of outperforming some of the most powerful language models in various scenarios. According to Anthropic, Claude 3.5 Sonnet has outperformed GPT-4o, Gemini 1.5 Pro, and Meta’s Llama 3 400B in nine general benchmarks, which measure the capabilities of each model.

Claude 3.5 Sonnet is more precise than GPT-4o when generating and analyzing code, according to Anthropic’s data. It also excels at transcribing text from images. Moreover, its language capabilities have been enhanced to provide more human-like responses, even incorporating touches of humor to make interactions more friendly.

Claude 3.5 Sonnet is now available to all users with access to Claude, and it can be accessed via the web or the iOS app. However, this isn’t the only new development from Anthropic.

Claude 3.5 Sonnet is not the only announcement

Anthropic has also introduced Artifacts, a feature designed to enhance the user experience with Claude 3.5 Sonnet and other versions of the model. According to Anthropic, Artifacts provide a preview of the creation process for anything a user requests. For instance, if someone asks Claude 3.5 Sonnet to generate an element using SVG, the model will display the code in a side view.

Although the company hasn’t confirmed specifics, it appears that both Claude 3.5 Sonnet and the Artifacts feature will be available for free. However, users will need to manually activate Artifacts through the chatbot settings.