Gemini 4 Argon API on FastMetal: 1M context, image input

Google's Gemini 4 Argon is available on FastMetal in chat, the OpenAI-compatible API and MCP, under the model id gemini-4-argon. Google announced it on September 30, 2026 as the first model of the Gemini 4 generation, a frontier model for real-world coding, enterprise knowledge work and cyber defense.

At launch on October 13 we offered Argon in chat only. On October 21, 2026 we moved it to a different upstream host, which put it on the API and MCP as well. That host does not accept PDF attachments, so Argon on FastMetal no longer reads PDFs. Paste the text of a PDF into your message instead.

What it is built for

• Long, multi-step work. Google's main claim for Argon is that it keeps deep reasoning going across long tasks with many steps.
• Software engineering. Google reports 77.9% on DeepSWE v1.1, which measures development work in real repositories, and calls it a new high. That is Google's own figure.
• Legal and financial documents. Google names enterprise knowledge work as a main use: reading a set of documents and turning them into a decision or a next step.
• A 1M-token context. We serve it with 1,048,576 tokens, enough for a long contract, a full specification or a whole set of documents in one request.
• A reasoning model that reads images. It thinks before it answers. We checked image input with real files before launch.

Pricing

Gemini 4 Argon is billed per million input tokens and per million output tokens. Thinking tokens count as output.

Our rate follows the introductory price Google announced for Argon. Google has said the price goes up when the introductory period ends, and our rate will change when it does. Current rates are on the model page.

Billing is prepaid: you load a balance by card, from $5, and you can never spend more than you loaded. See pricing.

How to use it

In chat:

1. Create a FastMetal account.
2. Load a balance.
3. Open chat and pick Gemini 4 Argon from the model list. Attach images to a message as you need them.

From the API, send gemini-4-argon as the model to the OpenAI-compatible endpoint:

curl https://api.fastmetal.ai/v1/chat/completions \
  -H "Authorization: Bearer $FASTMETAL_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gemini-4-argon",
    "messages": [{"role": "user", "content": "Summarize the three key points of this contract: ..."}]
  }'

Compare real answers first

The model page shows Argon's actual answers to our 43 shared prompts: summaries, bug fixes, translation and more. Compare them with Google's Gemini 3.7 Flash or with the model you use today. For Argon against its newer sibling, read Gemini 4 Carbon vs Gemini 4 Argon. Everything we serve is in the model catalog.