$2 Per Million Tokens: What Gemini 4 Argon Costs Indian Devs

Google says engineers using Gemini 4 Argon inside its Antigravity coding tool pushed agentic code submissions up by more than 35%, while cutting code rollbacks by nearly a third. That same model is now open to any paying developer in India.

The launch price is $2, roughly ₹190, per million input tokens. The catch is that this is an introductory rate, and it doubles after November 30.

Quick answer

• Gemini 4 Argon is Google's new flagship model, rolling out globally to developers through the paid Gemini API and to Google AI Pro and Ultra subscribers in the Gemini app and AI Mode in Search.
• Introductory API pricing is $2 (about ₹190) per million input tokens and $10 (about ₹950) per million output tokens.
• After November 30, prices move to $4 (about ₹380) input and $20 (about ₹1,900) output per million tokens.
• The Antigravity coding tool is limited to Ultra subscribers, and there is no word yet on free users.

Why this launch matters now

Google has had a rough year in AI. The company said in September that it no longer planned to release Gemini 3.5 Pro, a model CEO Sundar Pichai had originally said would arrive in June, according to The Economic Times.

It also overhauled DeepMind after co-founder and CEO Demis Hassabis stepped aside and several Gemini leaders left. The Verge called Argon Google's first major frontier release since Gemini 3 in late 2025.

Argon lands just a day after Meta announced its own frontier model, Muse Wonder Full. So for developers, the real question is not which lab is winning. It is which model gives the best work per rupee.

What Argon actually costs

Google is pricing Argon aggressively. The Verge reports the introductory price is around a fifth of what OpenAI's GPT-6 Astra and Anthropic's Fable 5 models cost.

Gemini 4 Argon pricing: $2 input and $10 output per million tokens now, rising to $4 and $20 after November 30

Source: Google, India Today

Here is the same pricing in rupees, using India Today's rough conversions:

Cached input tokens get a 95% discount on the input price, Google says. If your app sends the same long system prompt or document again and again, caching is where the real savings sit.

One warning from The Verge is worth repeating. A cheaper per-token price does not always mean a cheaper bill, because different models can use very different numbers of tokens for the same task.

What it can do, by Google's numbers

Google claims Argon beats GPT-6 Astra, Claude Fable 5.1 and Claude Opus 5.5 on benchmarks across coding, cybersecurity, science, finance and law. It also says Argon ranks first on the Vals Index, which tracks performance on finance, coding, legal and tax work.

India Today adds that Argon beats GPT-6 Astra and Claude Opus 5.5 on AutomationBench and the coding test DeepSWE 1.1.

The headline spec for builders is output length. Argon gets a 1 million token output limit, compared with roughly 1,28,000 for rival models, India Today reports. That matters for long code refactors and large reports.

Google's own engineering examples are specific:

• The Chrome team used Argon to separate core webpage handling from a roughly 6 million-line legacy folder inside Chromium's 40 million-line codebase, finishing months of work in weeks.
• The Agent Development Kit team says Argon cut in-memory session prep time by 40% in long sessions and sped up cold imports by 21%.
• Researchers made the xTB quantum chemistry package up to 2.3x faster at predicting molecular structures.
• Argon delivered a 2.85x inference speedup for Gemma, from about 37,900 to over 108,000 tokens per second.

On robotics, Google's developer blog says Argon scores 87% on the Embodied Reasoning (ERQA) benchmark, which tests spatial problem solving in 3D environments.

Remember that these are vendor claims. Treat them as a reason to test, not as a verdict.

What it means for Indian users and freshers

If you are a student or fresher building projects, the paid API tier is the entry point. At introductory rates, a small hobby app that sends modest prompts can be tested cheaply before November 30. Build your evaluation set now, while the price is low.

If you already pay for Google AI Pro or Ultra, you can try Argon in the Gemini app and in AI Mode in Search. India Today notes the Gemini app can connect to over 40 apps, including Adobe and Shopify, for tasks like ordering groceries or booking tables.

If you write code for a living, the Antigravity integration is the interesting part, but it is limited to Ultra subscribers for now.

If you work in security, Google is expanding its Fairwind Program to hundreds of additional partners. It gives vetted cybersecurity defenders a version of Argon without its cyber guardrails.

Google will also give 10,000 scientists six months of free access to Antigravity's highest usage tier, The Economic Times reports.

Mistakes to avoid

Budgeting on the intro price. Plan your costs on the after-November-30 rates of $4 and $20 per million tokens. Anything else will surprise you in December.

Ignoring output tokens. Output costs five times as much as input. Long answers, and the model's thinking, add up fast.

Skipping caching. A 95% cut on repeated input is large. Structure prompts so the fixed part comes first.

Trusting benchmarks blindly. Run Argon on ten of your own real tasks, next to the model you use today, and compare both quality and total tokens.

FAQs

Is Gemini 4 Argon available in India? Yes. Google says it is rolling out globally, for developers through the Gemini API and for Google AI Pro and Ultra subscribers in the Gemini app and AI Mode in Search.

How much does Gemini 4 Argon cost in rupees? India Today puts the introductory price at roughly ₹190 per million input tokens and ₹950 per million output tokens, rising to about ₹380 and ₹1,900 after November 30.

Can free Gemini users try Argon? Not yet. The Verge says it is not clear when, or if, Argon will reach Plus subscribers or free users.

Who were the early testers? The Economic Times lists Deloitte, LexisNexis, Ambience Healthcare and Isomorphic Labs among early customers.

Prices are Google's published rates; rupee figures are India Today's rough conversions. Check Google's official pricing page before you budget.

Sources

• The Economic Times: Google launches flagship AI model Gemini 4 Argon for developers, businesses
• The Verge: Google finally releases Gemini 4 Argon
• Google for Developers: Gemini 4 Argon for Robotics
• India Today: Google rolls out Gemini 4 Argon globally
• Google: Gemini 4 Argon, built to solve your hardest problems