Anthropic recently released a blog post, officially launched Claude Instant 1.2 AI model, and has provided API access to enterprises.
Anthropic says the Claude Instant 1.2 AI model is more powerful, less expensive and more secure than previous versions.
In terms of performance, the Claude Instant 1.2 AI model can handle up to 100,000 tokens, meaning it can process hundreds of pages of content at once, and output memos, letters, stories, and other text content containing thousands of tokens at once.
Anthropic says Claude Instant 1.2 generates responses that are longer, more structured, and follow formatting instructions better, with improved quote extraction, multilingual capabilities, question answering, and more.
The biggest highlights of Claude Instant 1.2 are a new top evaluation score for the Claude model, further improved security, reduced illusions, and greater resistance to jailbreaking.
Note: Illusion is when a model generates incorrect or meaningless text, and jailbreaking is a technique for bypassing security features set on large language models. For more information on the Claude Instant 1.2 AI model, you can visit here.
