Anthropic unveils improved Claude 2.1 AI model

The Anthropic team unveiled a new AI model, Claude 2.1, with an expanded context window of 200,000 tokens, equivalent to about 500 pages of text.

Our new model Claude 2.1 offers an industry-leading 200K token context window, a 2x decrease in hallucination rates, system prompts, tool use, and updated pricing.

Claude 2.1 is available over API in our Console, and is powering our https://t.co/uLbS2JNczH chat experience. pic.twitter.com/T1XdQreluH

— Anthropic (@AnthropicAI) November 21, 2023

According to the developers, users can feed the neural network large documents such as code excerpts or financial reports. Processing and executing a query takes ‘a few minutes’.

Claude 2.1 significantly reduces the rate of ‘hallucinations’: the number of false statements has halved versus the previous version.

The new model also features substantial improvements in understanding and generalisation, especially for long and complex materials requiring a high degree of precision.

Anthropic also added a beta feature of Claude-based tools that can be integrated with third-party applications to assist developers.

The tool’s capabilities include:

• using a calculator for complex computations;
• transforming textual queries into structured code;
• web search to answer the posed question;
• executing commands via API;
• configuring client services.

Claude 2.1 is available in the chat-bot interface only to users with a Claude Pro subscription.

Earlier in February, media reported that Anthropic had raised $300 million from Google. In March, the Spark Capital invested the same amount in the AI startup. At that time, its valuation reached $4.1 billion.

In August, Anthropic raised $100 million from the South Korean telecom giant SK Telecom. In September, Amazon announced strategic investments in the startup worth up to $4 billion. Later, Google pledged to allocate additional $2 billion to the project.