Anthropic says its new family of Claude 3 AI models can match or even surpass Gemini and ChatGPT, reports The Verge.

The company presented three new models: Claude 3 Haiku, Sonnet, and Opus. Among them, Opus is the largest and smartest model.

The company also says that the new model makes fewer mistakes in answers.

Compared to previous models, Claude 3 can answer more questions, understand longer queries, and be more accurate.

According to Anthropic’s own tests, the Claude 3 Opus outperforms the competition in areas ranging from math and coding to general knowledge and text reasoning.

Currently, Claude 3 Opus and Sonnet models are already available on claude.ai and its API, Haiku will be released later. The models can be used in chatbots, autocomplete, and data extraction tasks.

Anthropic was founded by former OpenAI employees. Last September 2023, the company received $4 billion in investment from Amazon, which is why it became one of the objects of the Federal Trade Commission’s investigation.