Anthropic is enhancing its AI coding model, Claude Sonnet 4, by increasing its context window to 1 million tokens, allowing up to 750,000 words in a single prompt. This update, aimed at attracting enterprise developers, significantly surpasses the previous limit of 200,000 tokens and even exceeds OpenAI’s GPT-5’s 400,000-token capacity. Available through cloud partners like Amazon Bedrock and Google Cloud’s Vertex AI, this expansion aligns with Anthropic’s strategy to dominate the AI model market, particularly for coding platforms like GitHub Copilot and Anysphere’s Cursor. Despite competition from GPT-5, which recently became the default model for Cursor, Anthropic’s product lead, Brad Abrams, remains optimistic about the growing API usage. The firm plans to adjust pricing for extended token use, reflecting ongoing efforts to improve AI performance and user experience. As AI models benefit from larger context, Claude’s enhancements are expected to boost its capability in complex coding tasks.
Source link
Share
Read more