Anthropic releases more powerful AI model Claude Instant 1.2

IT House News on August 12, Anthropic recently published a blog post, officially launching the Claude Instant 1.2 AI model, and has provided API access to enterprises.

Anthropic Company stated that the Claude Instant 1.2 AI model has stronger performance, lower price, and higher security than previous versions.

In terms of performance, the Claude Instant 1.2 AI model can handle up to 100,000 tokens, means that it can process hundreds of pages of content at the same time and output memos, letters, and stories containing thousands of tokens at one time and other text content.

Anthropic says Claude Instant 1.2 generates improved responses that are longer, more structured, and follow formatting instructions, as well as improved citation extraction, multilingual capabilities, and question answering

The biggest highlight of Claude Instant 1.2 is to improve the highest evaluation score of the Claude model, enhance security, reduce hallucinations, and more effectively prevent jailbreaks

IT House pointed out that hallucination refers to a model that generates incorrect or meaningless text, while jailbreaking is a technology that bypasses the security features set on large language models. To learn more about the Claude Instant 1.2 AI model, please visit this link

The above is the detailed content of Anthropic releases more powerful AI model Claude Instant 1.2. For more information, please follow other related articles on the PHP Chinese website!