Home » blog » Gemini 4 Argon Pricing: Half Price Until November 30, Then Rates Double

Gemini 4 Argon Pricing: Half Price Until November 30, Then Rates Double

Tram cong nghe AI blog

Gemini 4 Argon pricing starts at $2 per million input tokens and $10 per million output tokens, but only until November 30, 2026. Google made Gemini 4 Argon generally available on October 13, two weeks after it reserved the model for a small group of vetted cyber defenders. On December 1, both API rates double.

Key Takeaways

• General availability: October 13, 2026, on the paid tier of the Gemini API and in Google AI Studio (model ID gemini-4-argon).
• Introductory price (through Nov 30): $2 per 1M input tokens, $10 per 1M output tokens, $0.10 per 1M cached input tokens.
• Standard price (from Dec 1): $4 input, $20 output, $0.20 cached input. Every rate doubles.
• Vs. GPT-6 Astra: OpenAI charges $10/$50, five times Argon’s current rate.
• Consumers: Google AI Pro subscribers ($19.99/month) get Argon in the Gemini app and in AI Mode in Search.
• Not included: the free tier gets no access to Argon.

What Is Gemini 4 Argon?

Gemini 4 Argon is Google’s strongest model. Google’s release notes call it “our most capable model built for complex coding, agentic, math, and science tasks.” Google aims it at long software projects, enterprise knowledge work and cyber defense.

Until this week, Argon was limited to trusted security teams in Google’s Fairwind Program, Google’s own staff and the US government. Now it is open to paying developers and subscribers.

One notable spec: according to Google, Argon can write up to 1 million output tokens in a single response, up from 64,000 on earlier Gemini models. Output tokens are the text the model generates back to you.

Gemini 4 Argon Pricing

Here is the full API price table. “Cached input” means repeated input (like a long system prompt or document) that the API stores and re-reads at a much lower price.

Two details matter for your bill:

• Thinking tokens are billed as output. Argon “thinks” before it answers, and those hidden reasoning tokens are charged at the output rate. On hard tasks, this can make output the biggest part of your cost.
• Cached input is a 95% discount. If your app sends the same long context again and again, caching cuts input costs sharply.

Watch out: the low price is temporary. Any cost estimate you make today will double on December 1. Budget for the standard rate, not the launch rate.

Gemini 4 Argon vs GPT-6 Astra and Claude Opus 5.5 on price

At launch prices, Argon undercuts its closest rivals. The source article notes that at introductory rates Argon is half the price of Claude Opus 5.5. After the December increase, Argon lands in the same price band as Claude Opus 5.5 and Meta’s Muse Wonder, which launched a day earlier.

The strategy is clear: Google is starting low to win workloads, then raising prices once teams have built on the model.

Gemini 4 Argon Benchmarks

All numbers below come from Google’s own table, according to Google’s reported figures. No independent lab has reproduced them in the source we used.

Long-horizon coding (DeepSWE v1.1)

DeepSWE v1.1 is a benchmark built on real-world, long-running coding tasks.

Security and other coding tests

• CWE-bench v1 (vulnerability repair): Argon tied GPT-6 Astra at 68.0%.
• FrontierSWE v2 and Terminal-Bench 4.0: Argon trailed rivals, according to Google’s own table.

Google also says teams of Argon agents analyzed profiling data across its data centers and freed more than 300 TiB of memory.

How to Access Gemini 4 Argon

Who can use Gemini 4 Argon right now:

Google also says it will give 10,000 scientists free access to the top Antigravity usage tier for six months.

What you should do: if you run coding agents or heavy workloads, use the seven weeks before December 1 to test Argon on real work. Measure how many output and thinking tokens your tasks use, then multiply by the standard $4/$20 rate to see your real long-term cost.

Is Gemini 4 Argon Worth It?

Pros

• Very cheap for a frontier model until November 30: one fifth of GPT-6 Astra’s rate.
• Top score on DeepSWE v1.1 in Google’s table (77.9%).
• Up to 1 million output tokens in one response.
• Available to AI Pro subscribers in the Gemini app, not just developers.

Cons

• The price doubles on December 1, putting it in the same band as Claude Opus 5.5.
• Thinking tokens are billed as output, which can inflate costs.
• Benchmarks are Google’s own; Argon trailed on FrontierSWE v2 and Terminal-Bench 4.0.
• No free tier, and Antigravity access requires the $99.99 AI Ultra plan.

The verdict: the launch price is a good window to test Argon cheaply. Decide whether to keep it based on the standard price, not the promo.

Gemini 4 Argon FAQ

How much does Gemini 4 Argon cost?
Until November 30, 2026, Gemini 4 Argon costs $2 per million input tokens and $10 per million output tokens. From December 1, it costs $4 and $20.

When does Gemini 4 Argon pricing go up?
On December 1, 2026. Input, output and cached input rates all double.

Is Gemini 4 Argon free?
No. The free tier of the Gemini API gets no access. Consumers need a Google AI Pro subscription to use it in the Gemini app.

Is Gemini 4 Argon cheaper than GPT-6 Astra?
Yes. GPT-6 Astra costs $10/$50 per million tokens, five times Argon’s current rate.

How do I use Gemini 4 Argon?
Developers can call gemini-4-argon on the paid Gemini API or in Google AI Studio with a paid API key. AI Pro subscribers can use it in the Gemini app and AI Mode in Search.

Sources

• Google Opens Gemini 4 Argon at $2/$10 Pricing – Technology.org

Last updated: October 14, 2026