ChatGPT competitor Claude has announced it has launched its Model Clude 2 with enhanced performance and extended responses, accessible via API.
User feedback highlights Claude’s user-friendly nature, reduced likelihood of generating harmful outputs, and improved memory. Claude 2 has shown significant improvements in areas such as coding, mathematics, and reasoning, scoring 71.2% on the Codex HumanEval, a Python coding test, and 76.5% on the multiple choice section of the Bar exam.
The model has been fortified against jailbreaks and is less likely to produce offensive or dangerous outputs. In internal red-teaming evaluations, Claude 2 was twice as effective at generating harmless responses compared to its predecessor, Claude 1.3. Thousands of businesses are now using Claude, with partners like Jasper and Sourcegraph collaborating to develop new AI applications that prioritize safety and reliability.
The company behind Claude, Anthropic, is an American artificial intelligence (AI) startup and public-benefit corporation, was founded in 2021 by former members of OpenAI.