Google Releases Gemini 4 Argon with 1M Token Output Cap: what are the key facts?

On October 14, Google opened the Gemini 4 Argon model to developers and enterprises worldwide.

Google Releases Gemini 4 Argon with 1M Token Output Cap: what are the key facts?

The model supports output of up to roughly 1 million tokens, sharing the same context window as input reasoning.

Google Releases Gemini 4 Argon with 1M Token Output Cap: what are the key facts?

Limited-time introductory pricing is $2 per million input tokens and $10 per million output tokens.

Google Releases Gemini 4 Argon with 1M Token Output Cap: what are the key facts?

After the promotion ends on November 30, the output price rises to $20 per million tokens.

Google Releases Gemini 4 Argon with 1M Token Output Cap: what are the key facts?

The model ranks first in the Vals Index performance ranking for the economic domain.

Google Releases Gemini 4 Argon with 1M Token Output Cap: what are the key facts?

Google is giving 10,000 scientists six months of free access at the highest tier.

Physix Frontier · News Briefing Card (enbrief · Oct 14, 2026)

Google Releases Gemini 4 Argon with 1M Token Output Cap

KEY FACTS

On October 14, Google opened the Gemini 4 Argon model to developers and enterprises worldwide.
The model supports output of up to roughly 1 million tokens, sharing the same context window as input reasoning.
Limited-time introductory pricing is $2 per million input tokens and $10 per million output tokens.
After the promotion ends on November 30, the output price rises to $20 per million tokens.
The model ranks first in the Vals Index performance ranking for the economic domain.
Google is giving 10,000 scientists six months of free access at the highest tier.

KEY DATA

1,048,576Max output tokens

$2 per million tokensIntroductory input price

$10 per million tokensIntroductory output price

nearly halfViolation rate reduction

PHYSIX OBSERVATION

A million-token output cap makes one-shot generation of long documents and large codebases a reality, but because input and output share a single context window, the actual usable space gets squeezed by the reasoning process. The introductory price is half the post-promotion price, an obvious limited-time hook to grab developers. The real signal is that Google is using low pricing plus free research slots to push the model into law, finance, and research - the scenarios with the strongest willingness to pay - directly targeting OpenAI's enterprise base.

Source: enbrief original report ↗