Anthropic Expands Claude's Memory to 1 Million Tokens, Offers AI to All US Government Branches for $1

Hans India

3 days ago

Anthropic has unveiled two significant developments for its Claude AI models — a massive leap in processing capacity and an unprecedented public sector offer. The company is opening access to all three branches of the US government — executive, legislative, and judiciary — for just $1 per year, marking a bold expansion beyond its competitors' outreach.
The move closely follows OpenAI's recent announcement to provide ChatGPT Enterprise to the entire federal executive branch for the same nominal fee. But Anthropic is taking it a step further by targeting every branch of the government, with the goal of embedding Claude into a wider range of public sector operations. The offer will be valid for one year.
Claude Gets a Memory Boost
Alongside the government-focused push, Anthropic is rolling out a fivefold increase in the processing capacity of its Claude Sonnet 4 model. The AI's context window — the amount of information it can process in one go — now stands at 1 million tokens, equivalent to roughly 750,000 words or 75,000 lines of code. To visualize it, that's more than the entire Lord of the Rings trilogy in a single prompt.
This is a substantial jump from Claude's previous 200,000-token limit and surpasses the 400,000-token capacity of OpenAI's GPT-5. The expanded context window will be available to enterprise clients via Anthropic's API and through cloud partners such as Amazon Bedrock and Google Cloud's Vertex AI, making it easier for existing customers to integrate.
Government-Specific AI with High Security
Anthropic will provide two tailored versions of its AI — Claude for Enterprise and Claude for Government. The latter is designed to meet FedRAMP High compliance, the strictest security standard for handling sensitive but unclassified federal data. This ensures federal employees can safely use the AI for tasks involving confidential materials.
Potential applications range from summarizing lengthy reports and translating documents into multiple languages, to assisting in complex technical projects. The company also plans to support agencies in integrating Claude into their systems without requiring major infrastructure overhauls, thanks to partnerships with AWS, Google Cloud, and Palantir.
Part of a Larger AI Race
The announcement underscores the intensifying competition between AI companies for a foothold in the government sector. Recently, the US General Services Administration approved Anthropic, OpenAI, and Google DeepMind as official AI vendors for civilian agencies, streamlining the procurement process.
Anthropic already collaborates with the US Department of Defense and other agencies on national security-related AI projects. However, the new initiative signals a broader ambition to support government functions in areas such as scientific research, public health, and citizen services.
With OpenAI's executive branch deal already in motion, industry watchers are now waiting to see whether Google or other AI leaders will counter with offers of their own. For now, Anthropic's latest upgrade and $1 deal mark a significant statement of intent — both in technological advancement and in the high-stakes race to shape the role of AI in public governance.