Gemini 4 Argon rumored specs: 1M token output, LVBench 91.7%

Athrix ☄️ @athrix_codesOriginally posted 2026.10.09

Open original on X

Can't believe, last night without Gemini 4 Argon... Fr excited to build with Argon 😳 • 1 Million Token Output Limit • LVBench (91.7%) • DeepSWE v1.1 (77.9%) • AutomationBench (51.3%) • Ultra-Low Cost Caching And it is the lowest hallucination rate model in the world absolutely cooking every model on the benchmark