News

Anthropic Expands Claude’s Context Window to 1 Million Tokens for Enterprise AI

Cansin Cengiz

12 Aug 2025 — 2 min read

Claude AI Now Accepts 1 Million Token Prompts: What This Means for Enterprise and Developers

Anthropic has announced a significant upgrade to its Claude AI platform, allowing enterprise customers to send much longer prompts — up to one million tokens per request. This update is a major leap for AI usability, especially for businesses and developers working with complex datasets or extensive codebases.

What’s Changed?

Claude Sonnet 4, Anthropic’s popular AI model, now offers a context window that can process up to one million tokens. To put this in perspective:

That’s about 750,000 words — more than the entire Lord of the Rings trilogy combined.
Or roughly 75,000 lines of code in a single input.
This is five times larger than Claude’s previous limit (200,000 tokens), and more than double the 400,000 tokens currently offered by OpenAI’s GPT-5.

This expanded capability is available to API customers, as well as users accessing Claude through major cloud partners such as Amazon Bedrock and Google Cloud Vertex AI.

Why It Matters for Business and Developers

With a larger context window, Claude can now "see" entire projects, documents, or codebases at once. This has profound benefits for:

Software Engineering: Developers can prompt the AI with their entire code repository, making it easier for Claude to provide contextually accurate suggestions, refactorings, or feature implementations.
Enterprise Data Analysis: Large documents or datasets can be processed in a single prompt, improving efficiency for research, compliance, and strategic planning.
Long-Form Content Generation: Businesses can generate or analyze lengthy reports, proposals, or knowledge bases without splitting them into smaller parts.

Brad Abrams, Anthropic’s product lead for Claude, emphasized that this upgrade should bring “a lot of benefit” to AI-powered coding platforms and other enterprise applications, particularly by enhancing performance on long, complex tasks.

How Does Claude Compare to Competitors?

Anthropic’s move comes as the competition among AI model providers heats up:

OpenAI’s GPT-5 offers a 400,000 token context window.
Google’s Gemini 2.5 Pro boasts a 2 million token limit.
Meta’s Llama 4 Scout advertises a staggering 10 million token window.

However, not all models can process large contexts with the same effectiveness. Abrams noted that Anthropic’s research team focused on increasing not just the raw context window, but the "effective context window" — meaning Claude is designed to comprehend and utilize most of the input data, not just store it.

Pricing Updates

With the new capabilities, Anthropic is adjusting its API pricing for larger prompts. For requests over 200,000 tokens:

Input tokens: $6 per million
Output tokens: $22.50 per million

For comparison, previous rates were $3 per million input tokens and $15 per million output tokens.

Driven by Enterprise Demand

Anthropic has built a strong reputation in the enterprise sector, with customers including AI coding platforms like GitHub Copilot, Windsurf, and Anysphere’s Cursor. This focus on API and enterprise sales differentiates Anthropic from competitors like OpenAI, which generates most of its revenue from consumer subscriptions.

Looking Ahead

Anthropic continues to enhance Claude’s capabilities, recently releasing an updated version of its largest model, Claude Opus 4.1, to further improve coding and reasoning tasks. As context windows grow, the potential for AI to tackle even more complex, long-horizon problems increases — but so do the challenges in processing and understanding massive inputs.