Claude 2.1 released with a 200K token context window and lower model hallucination rates

Anthropic has introduced Claude 2.1, an updated version of their AI chat system, featuring a unique 200K token context window, decreased model hallucination rates, system prompts, and a beta feature called 'tool use'.

The 200K token context window allows users to input up to 150,000 words or more than 500 pages of content, enhancing Claude's ability to summarize, answer questions, predict trends, and compare documents. This feature also reduces task completion time.

The new version also boasts a 2x reduction in false statements, enhancing its credibility and reliability for businesses. The model's improved ability to avoid incorrect information is backed by rigorous testing. Claude 2.1 also shows significant improvements in understanding and summarizing lengthy, complex documents.

The new version is available through the API for both free and Pro users on claude.ai, with the 200K token context window feature exclusive to Pro users.