Anthropic, a leading AI startup, today announced that its popular Claude Sonnet 4 LLM now supports up to 1 million tokens of context. Following Google Gemini, Anthropic is now the second major model provider to offer 1 million context tokens.
Previously, the Claude Sonnet 4 API supported only 200K context tokens. This fivefold increase will allow developers to send entire codebases that are over 75,000 lines of code in a single request.
The expanded context support is now in public beta on the Anthropic API and via Amazon Bedrock, with availability on Google Cloud’s Vertex AI coming soon. However, long-context support is currently limited to Tier 4 developers with custom rate limits. Anthropic noted that it will expand access to more developers in the coming weeks.
Since the larger token windows require more compute power, Anthropic is introducing special pricing. For prompts under 200K tokens, Sonnet 4 will cost $3 per million input tokens and $15 per million output tokens. For prompts exceeding 200K tokens, the cost will be $6 per million input tokens and $22.50 per million output tokens.
Developers can reduce costs by using prompt caching and batch processing. For example, batch processing can provide a 50% discount on the 1M context window pricing.
During a recent AMA session on Reddit, OpenAI leaders discussed supporting a long context window for their models. OpenAI CEO Sam Altman mentioned that OpenAI has not seen significant demand for long context lengths from its users but is open to supporting it if there is sufficient interest. Since they are compute-constrained, they want to focus on other priorities.
Michelle Pokrass from the OpenAI team wrote that they would have liked to offer longer context, up to 1 million tokens in GPT-5, particularly for API use cases, but due to high GPU demand, they did not.
Anthropic"s 1M context support places it in direct competition with Google Gemini for long-context capabilities, putting pressure on OpenAI to reconsider its roadmap.