Gemini 4 Argon API launch: pricing and usage

• Published: October 14, 2026
• Last verified: October 14, 2026
• Status: Active; available in the default token group

​

Key facts

​

Pricing

• Google’s announcement says the price rises to $4 input and $20 output after the launch period, without giving an end date. LaoZhang API will post an announcement when its price changes.
• Thinking tokens are billed at the output rate. For example, a request with 53 input tokens, 6 answer tokens, and 832 thinking tokens is billed for 838 output tokens, about $0.0085.
• For current rates, see the model and pricing catalog and the console pricing page.

​

Call Gemini 4 Argon

export LAOZHANG_API_KEY="replace-with-your-LaoZhang-API-key"

curl https://api.laozhang.ai/v1/chat/completions \
  -H "Authorization: Bearer $LAOZHANG_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gemini-4-argon",
    "reasoning_effort": "high",
    "messages": [
      {"role": "user", "content": "Explain a database index in one sentence."}
    ]
  }'

curl "https://api.laozhang.ai/v1beta/models/gemini-4-argon:generateContent" \
  -H "Authorization: Bearer $LAOZHANG_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "contents": [
      {"role": "user", "parts": [{"text": "Explain a database index in one sentence."}]}
    ],
    "generationConfig": {
      "thinkingConfig": {"thinkingLevel": "high"}
    }
  }'

import os
from openai import OpenAI

client = OpenAI(
    api_key=os.environ["LAOZHANG_API_KEY"],
    base_url="https://api.laozhang.ai/v1",
    timeout=600.0,
    max_retries=0,
)
response = client.chat.completions.create(
    model="gemini-4-argon",
    reasoning_effort="high",
    messages=[{"role": "user", "content": "Explain a database index in one sentence."}],
)
print(response.choices[0].message.content)

• In the OpenAI-compatible response, the reply is in choices[0].message.content. In the Gemini-native response, it’s in candidates[0].content.parts[].text.
• Token usage and the charge for each request are in call logs.
• For images, streaming, and multi-turn requests in the native format, see the Gemini protocol.

​

Thinking levels

• On the Gemini-native endpoint, minimal returns a 400.
• On the OpenAI-compatible endpoint, none returns an error that asks for a thinking level instead. Use low when you don’t need deep reasoning.
• Higher levels produce more thinking tokens, so they cost more and take longer.

​

Who is affected

• Affected: projects that call Gemini through LaoZhang API for complex coding, long-document analysis, or multi-step reasoning. An API key in the default group with the Usage first or Usage-based billing mode can call it.
• Not affected: existing models. gemini-3.8-flash and the other Gemini models remain available, and current integrations need no changes.

​

Frequently asked questions

​

Do I need a Google account or early access?

​

Will the price change after the launch period?

​

Should I use Gemini 4 Argon or Gemini 3.8 Flash?

​

Why do minimal and none return an error?

​

What if a long request times out?

• OpenAI-compatible endpoint: add "stream": true to the request.
• Gemini-native endpoint: switch to streamGenerateContent?alt=sse.

​

Sources and related links

• Google: Gemini 4 Argon announcement: release date, positioning, output limit, launch and post-launch prices, and rollout plan
• Model and pricing catalog: LaoZhang API’s current rates
• Console pricing page: the rates that apply to your account
• Text generation API: protocols and models
• Gemini protocol: generateContent
• Gemini 3.8 Flash launch
• Latest announcements