Google Gemini 4 Argon Beats Claude Opus 5.5 on Arena AI

Google's Gemini 4 Argon generates up to 1 million tokens per response, costs from $2 per million input tokens and $10 per million output tokens, and cuts hallucination rates to 15%. The model ranks first in workflow orchestration and agentic task benchmarks.

• Up to 1 million tokens in a single response without losing context
• Pricing starts at $2 per 1M input and $10 per 1M output tokens
• Hallucination rate cut to 15% versus 50%+ for rivals
• Ranks first in workflow orchestration and task delegation benchmarks