Google releases Gemini 4 Argon, a frontier model for complex tasks
Google released Gemini 4 Argon, its latest frontier model, available globally via the API in AI Studio and the Gemini platform, and to Google AI Ultra and Pro subscribers in the Gemini app and AI Mode in Search.
Key facts
- Introductory pricing is $2 per million input tokens and $10 per million output tokens, with cached input tokens at 95% off.
- The model supports a 1M unified input-output token limit via the Gemini API.
- Argon ranks #1 on the Vals Index, which measures performance across finance, coding, legal, and tax work.
- It scores 87% on the Embodied Reasoning (ERQA) benchmark and leads the Gray Swan Indirect Prompt Injection benchmark at 0.7% attack success.
- On private Gamow Labs evals, Argon scores 57% on RareBench and 52.2% on LabBench.
Why it matters
Google says Argon outperforms other frontier models on benchmarks across domains and has already been used internally for large codebase refactors, quantum computing work, and enzyme design. It also expands the company's safety stack with agent alignment monitoring and prompt injection defenses.