Anthropic Removes Surcharge: 1 Million Token Context Window Now at Standard Pricing

Anthropic has quietly made one of the most significant API pricing changes of the year. As of this week, the 1 million token context window for Claude Opus 4.6 and Claude Sonnet 4.6 is available at standard per-token rates — with no premium surcharge.

Previously, using the full context window incurred significantly higher costs. Now developers and enterprises pay exactly the same rate whether they send 10,000 or 1,000,000 tokens in a single prompt.

What This Means in Practice

A 1 million token context window is equivalent to roughly 750,000 words — enough to load an entire codebase, a complete legal document archive, or a novel and its references in a single request. Without the surcharge, this opens up entirely new use cases that were previously too expensive:

• Full codebase analysis: Load entire repositories and ask about architecture, dependencies, and bugs
• Large-scale document review: Contracts, reports, and correspondence analyzed in one request
• Long-term conversation history: Applications that need to "remember" months of context
• RAG replacement: Some use cases can now skip complex RAG architecture entirely

Competitive Landscape

This change positions Anthropic ahead of most competitors on context-to-cost ratio. GPT-5.4 recently launched with a 1 million token context window, but OpenAI's pricing remains more complex. Gemini 3.1 Pro has long offered 1 million token context, but Anthropic users now have equivalent capacity without penalty.

For CIOs and development teams working with large datasets and complex workflows, this represents a real cost saving and a simplification of architecture decisions.