Gemini 4 Argon
Gemini 4 Argon is Google's stable October 2026 model for text, image, video, audio and PDF input. It generates text.
What the release provides
Its 1,048,576-token limit combines input and output. Synchronous output defaults to 65,536 tokens; long decoding can use the shared window. Thinking supports low, medium and high. Caching is implicit; explicit cache objects are not supported.Gemini 4 Argon official API documentation
How to use it well
Practical guidance. Budget the source material and the requested answer together. A long report can consume the room needed for its own response. Compare completed tasks at each thinking level, including tool results, instead of judging only the first token.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Access | API |
| Model identifiers | gemini-4-argon |
| Context window | 1,048,576 tokens Unified input and output window. Synchronous output defaults to 65,536 tokens; long decoding uses the remaining shared window. Gemini 4 Argon official API documentation |