Google Ships Gemini 4 Argon Globally, Betting the Frontier Race on Long-Context Agents
Google DeepMind made Gemini 4 Argon available worldwide to developers and enterprises, pairing a 1M-token context window with agent-focused benchmarks and aggressive $2-per-million-token input pricing.
Google re-entered the frontier model conversation on Wednesday with the global launch of Gemini 4 Argon, available immediately to developers and enterprises on the paid Gemini API. As @DanielZambrini noted, the rollout was global from day one — not the usual staggered regional drip — signaling that Google wants this model in production workflows fast.
The headline specs lean hard into the agentic era. Argon ships with a 1M-token context window and benchmarks positioned around coding and multi-step agent tasks rather than raw conversational polish. That framing matters: the competitive battleground has shifted from who sounds smartest in a chat window to who can hold state across a long, messy, tool-using session without losing the thread.
Get our free daily newsletter
Get this article free — plus the lead story every day — delivered to your inbox.
Want every article and the full archive? Upgrade anytime.
No spam. Unsubscribe anytime.