Google Releases Gemini 4 Argon with 1M-Token Output Capacity, Significantly Expanding AI Model Generation Limits

Google announced Gemini 4 Argon on September 30, 2026, a new frontier AI model that supports generating up to one million tokens in a single response—a capacity nearly 700% larger than the 128,000-token output limits of leading competitors from OpenAI and Anthropic. The 1 million-token output window represents a 16-fold increase from Gemini's previous 64,000-token ceiling, enabling substantially longer sequences of reasoning and code generation without interruption.