Anthropic has launched their new artificial intelligence model Claude 2.1 which will power their generative AI service Claude.ai. After many users requested support for processing longer data, Claude 2.1 now supports the processing of 200,000 context tokens or the equivalent of a 500-page prom (150,000 words). This is far superior to the capabilities of GPT-4 Turbo which only supports 128,000 context tokens/300 page proms.
Claude 2.1 also guarantees that their rate of hallucinations is twice as low in giving wrong answers as compared to the Claude 2.0 model. This indirectly guarantees to the user that the response received is more accurate.
Support for user processes, products and APIs is also offered in Beta form at this time after being one of the most requested features. This provides the ability to answer questions based on a database or perform web searches through an API. Instructing Claude 2.1 to use a calculator to perform complex calculations is also possible through this Beta feature.
This new update is available to all free users, but processing up to 200,000 words is limited to Claude Pro customers only.
Tags
APPS & GAMES