Anthropic unveils Claude 2.1 update with enhanced token limit and new features
![](https://cdn.prod.website-files.com/64a73265ea06a02bc3eab456/6561ffcab0669b030c0fdf27_claude.jpeg)
Previously 128,000 tokens, Claude’s context window has now been enlarged to 200,000 tokens
Anthropic has introduced a notable chatbot update. Anthropic's latest update, Claude 2.1, offers Pro tier users the ability to process up to 200,000 tokens simultaneously, equivalent to more than 500 pages of content. The update also promises reduced instances of hallucinations or falsehoods, enhanced functionalities such as web searches and calculator usage, support for custom, persistent instructions, and a new test window for prompt experimentation. This 200,000-token capacity sets a groundbreaking standard, exclusively available on Anthropic's Pro plan, surpassing the 32,000-token limit of the most expensive GPT-4 version, enabling in-depth analysis of extensive codebases or literary works.
Additionally, Anthropic has bridged the gap between Claude and ChatGPT by introducing a beta tool feature that allows users to connect API tools. Claude can now select the most appropriate tool based on contextual cues, including web searches and calculations, with users able to request specific API calls using natural language. Anthropic has further updated its developer console to include a test window for experimenting with new prompts and the capability to provide Claude with custom, persistent instructions, similar to the customization available with GPT-4, enabling tailored responses and personalities.