On 2026‑10‑13 I announced the launch of Gemini 4 Argon on X, highlighting massive performance gains, multimodal demos, and rollout details across paid tiers and internal testing.
Argon helped Gemma break 120,000+ tokens/sec on a single A100 on vLLM 🚀
Generating 64 live Python simulations at once 👇 https://t.co/gK8o1ZSaP7
It wrote a multi-pass canvas renderer that starts with broad background strokes, blocks in the buildings and river, and finishes with fine specular highlights on the water. https://t.co/Sj2PVssRnV
Gets over 140fps on M3 Ultra. What I'm doing now is having Argon build out generalized bindings for more games, to see if this could be applied to get fully native games on macOS without developers need
Gemini 4 Argon is the first model where it felt like an image, video and language model all at once.
A months‑long Chrome refactor at ~4x the usual pace
35%+ more agentic code submissions
Improved session creation in ADK by 40%
2.9x the throughput increases in SGLang for Gemma 4
The rollout is live in the Gemini API and Google AI Studio, with paid Pro and Ultra tiers on the Gemini App and Google Search (AI Mode). Internal testing involved over 100,000 Google employees.