Anthropic has introduced Claude 3, its latest artificial intelligence (AI) suite featuring the models Claude 3 Opus, Sonnet, and Haiku, boasting significant enhancements over its predecessors, the company announced in a blog post on Monday.
The standout model, Claude 3 Opus, surpasses GPT-4 and Google’s Gemini Ultra in benchmarks such as knowledge and reasoning levels and mathematical capabilities.
“In our quest to have a highly harmless model, Claude 2 would sometimes over-refuse,” Anthropic co-founder Daniela Amodei told CNBC in an interview. “When somebody would kind of bump up against some of the spicier topics or the trust and safety guardrails, sometimes Claude 2 would trend a little bit conservative in responding to those questions.”
For the first time, Anthropic integrates multimodal support, enabling users to analyze various unstructured data types, including photos and documents. While Sonnet and Haiku offer more affordable, compact options, Opus and Sonnet are now accessible in 159 countries, with Haiku’s release pending. This advancement marks a significant leap from Anthropic’s earlier stages, showcasing its rapid evolution into a key player in the generative AI space, with capabilities like summarizing texts up to 150,000 words and improved risk assessment in responses.
“Of course no model is perfect, and I think that’s a very important thing to say upfront,” said Amodei. “We’ve tried very diligently to make these models the intersection of as capable and as safe as possible. Of course there are going to be places where the model still makes something up from time to time.”
