Archive — US — techcrunch
Anthropic’s Claude AI model can now handle longer prompts
Anthropic has expanded the context window for its Claude Sonnet 4 AI model to one million tokens, allowing for up to 750,000 words in a single prompt. This enhancement aims to attract more enterprise customers amid competition from OpenAI's GPT-5, which offers a smaller context window. The update is expected to benefit AI coding platforms significantly, although higher charges apply for extensive token usage.
Key points
- Anthropic increased Claude's token limit to one million, enhancing its appeal to enterprise customers.
- Claude now handles requests up to 750,000 words, significantly exceeding previous limits (200,000 tokens) and rival GPT-5 (400,000 tokens).
- Longer context improves performance for complex coding tasks, especially in software engineering.
- Competitors like Google (2 million token limit) and Meta (10 million token limit) are pushing the boundaries of context windows.
- Claude's usage fees increase with prompts over 200,000 tokens: $6 per million input and $22.50 per million output tokens.
Why it matters
- AI trends indicate that larger context windows enhance model performance in coding and other complex tasks.
- The competitive landscape among AI developers is heating up, with companies like OpenAI and Google enhancing their models.
- In 2025, the emphasis on effective AI capabilities will grow, coinciding with industry events such as TechCrunch Disrupt.
- Anthropic's strategy reflects a broader shift towards enterprise solutions as companies seek powerful tools for software development.
Anthropic's expansion of Claude's context window positions it strongly against competitors like GPT-5, primarily serving enterprise coding needs while adapting its pricing. This advancement is likely to influence industry dynamics as AI capabilities evolve.