Google Gemini 4 Argon closes the gap with OpenAI and Anthropic but doesn't take a clear lead
Gemini 4 Argon matches GPT-6 Astra on benchmarks but burns 2x the tokens per task—Google closes the gap, but efficiency is the new moat.

Why it matters
Google's first frontier model in seven months reaches parity with OpenAI's best on independent tests but trails Anthropic's Claude Opus 5.5. Token efficiency—not just capability—now determines cost-of-ownership for practitioners; Argon's lower per-token price is offset by higher consumption, shifting the economic calculus for enterprise deployments.
The key facts
5 to knowGemini 4 Argon matches GPT-6 Astra in independent testing
Trails Anthropic Claude Opus 5.5 in capability
Argon consumes 2x+ tokens per task vs Astra despite lower per-token price
Select testers access first; API and paid tiers follow later
First Google frontier model release in over seven months
Go to the source
The Decoderthe-decoder.com
Publisher excerpt: Gemini 4 Argon is Google's first frontier model in over seven months. It matches GPT-6 Astra in independent testing but can't keep up with Anthropic's Claude Opus 5.5. The per-token price is low, but Argon burns through more than twice as many tokens per task as Astra. Select testers get access…