Anthropic Launches Claude 2 With Expanded Context and Better Coding Skills
Anthropic releases Claude 2, an upgraded AI chatbot offering a 100,000-token context window and enhanced coding abilities. The updated model is currently available in beta for users in the US and UK.
Anthropic releases Claude 2, an upgraded version of its AI chatbot that rivals ChatGPT with greatly improved safety and coding capabilities. The new model is available as a beta chat experience for users in the United States and the United Kingdom, while businesses access it through an API at the same price as the previous version. Claude 2 acts as a friendly assistant that helps with research, summarization, poetry, analysis, and coding tasks using natural language.
The updated chatbot features a significantly expanded context window of 100,000 tokens, allowing it to process approximately 75,000 words at once. This massive upgrade means users can submit entire large documents for the AI to ingest, analyze, and summarize. Additionally, Claude 2 generates longer coherent responses of about 3,000 words, a massive increase from the previous 400-word limit, enabling it to write entire memos, letters, and stories in a single prompt.
Claude 2 also shows remarkable improvements in coding, math, and reasoning, scoring higher on relevant benchmarks than its predecessor. It produces correctly formatted code in structured data languages such as JSON, XML, and YAML, making it highly useful for data transformation tasks. Despite these advancements and updated training data from early 2023, Anthropic warns that the chatbot still experiences confabulations or hallucinations and may produce errors on unfamiliar topics.