NOUMENOpen app
NOUMEN

Google · Frontier

Gemini 4 Argon

Added Oct 13, 2026

Compare models

Choose up to 3 models to compare.1 of 3 chosen

Loading models…

Use in the app

Google's Gemini 4 Argon is a flagship reasoning model for complex software engineering, mathematical and scientific problem solving, and multi-step agents. It accepts text, images, audio, video, and PDFs, with tool calling and structured outputs. Its 1,048,576-token context is shared across input, thinking, and output; synchronous output is capped at 65,536 tokens.

  • Reasoning
  • Tools
  • Vision
  • Structured output
  • PDF input
  • Audio input
  • Video input
Input
Text, Images, Video, Audio
Output
Text
Reasoning effort
low · medium · high
Discovery tier
Frontier

Pricing

USD per 1M tokens
Input
$2.00
Output
$10.00
Cache read
$0.10
Cache write
$0.19

Reference prices as published by the model provider. API example

Specifications

Context window
1M
Max output
65.5K
Average output (7 days)
1.6K tokens
Longer than 87% of 473 models

Side by side

Gemini 4 Argon compared

Specifications, prices, speed and benchmark scores next to each other. Benchmarks below show every compared model too.

Compare models

Choose up to 3 models to compare.1 of 3 chosen

Loading models…

Gemini 4 ArgonGoogleThis modelKimi K3.1 ProMoonshotRemove
ProviderGoogleMoonshot
Discovery tierFrontierFrontier
AddedOct 13, 2026Oct 24, 2026
Limits
Context window1.05M1.05M
Max output65.54K—
Average output1.6K tokens921 tokens
Pricing, USD per 1M tokens
Input$2.00$2.00
Output$10.00$10.00
Cache read$0.10Best$0.20
Cache write$0.19Best$2.00
Performance
Throughput113.7 tok/s114.4 tok/sBest
Time to first token10.0 s8.6 sBest
Request success (24h)98.5%Best82.9%
Benchmark indexes
Intelligence Index52.655.1Best
Capabilities
ReasoningYesYes
ToolsYesYes
VisionYesYes
Structured outputYesYes
PDF inputYesNo
Audio inputYesNo
Video inputYesYes
Reasoning effortlow, medium, highlow, high, max

Public evaluations

Benchmarks

Independent evaluation results, with how each score ranks among the models in the Noumen catalog.

  • Gemini 4 Argon
  • Kimi K3.1 Pro

Intelligence Index

  • Gemini 4 Argon52.6
  • Kimi K3.1 Pro55.1

Gemini 4 Argon tested at high reasoning effort

Gemini 4 Argon: Better than 97% of 326 scored models

3.8Catalog models by score55.1

Agentic work

  • AutomationBench-AA

    Workflow automation, penalized for guardrail violations

    • Gemini 4 Argon: 77.5%Top
    • Kimi K3.1 Pro: 73.3%

  • AutomationBench-AA tasks completed

    Workflows finished end to end without a guardrail violation

    • Gemini 4 Argon: 53.1%Top
    • Kimi K3.1 Pro: 46.0%

  • Harvey LAB-AAv1.1

    Legal tasks passed in full, with no material hallucination

    • Gemini 4 Argon: 3.9%
    • Kimi K3.1 Pro: —

  • AA-Briefcasev1.1

    Agentic knowledge work, rated head to head

    • Gemini 4 Argon: 1,486 Elo
    • Kimi K3.1 Pro: 1,678 EloTop

  • GDPval-AAv2.1

    Economically valuable occupational tasks, rated head to head

    • Gemini 4 Argon: 1,624 Elo
    • Kimi K3.1 Pro: 1,846 EloTop

Document reasoning

  • GDP.pdf

    Professional reasoning over PDF documents, all-pass rate

    • Gemini 4 Argon: 21.8%
    • Kimi K3.1 Pro: 31.4%Top

  • AA-LCRv1.1

    Reasoning across long documents

    • Gemini 4 Argon: 79.7%
    • Kimi K3.1 Pro: 86.3%Top

  • MLCR-AA

    Long-context reasoning over medical records

    • Gemini 4 Argon: 13.9%
    • Kimi K3.1 Pro: —

Reasoning and knowledge

  • Humanity’s Last Exam

    Expert-level questions across academic fields

    • Gemini 4 Argon: 57.1%Top
    • Kimi K3.1 Pro: 52.5%

Coding

  • Terminal-Benchv4.0

    Practical software engineering in a terminal

    • Gemini 4 Argon: 57.1%
    • Kimi K3.1 Pro: 67.2%Top

  • SciCode

    Scientific computing in Python

    • Gemini 4 Argon: 61.8%Top
    • Kimi K3.1 Pro: 60.8%

Legacy evaluations

  • AA-LCR (unversioned)

    Long-context reasoning, earlier grading

    • Gemini 4 Argon: 79.7%
    • Kimi K3.1 Pro: 86.3%Top

Scores from Artificial Analysis, updated Oct 26, 2026. Gemini 4 Argon was tested at high reasoning effort; compare scores measured at the same effort.Artificial Analysis

Live routing data

Performance and reliability

Recent averages across requests routed for this model. Your own speed depends on prompt length, reasoning effort and load.

Throughput
113.7 tok/s
Faster than 67% of 257 models
Time to first token
10.0 s
Quicker than 9% of 257 models
Prompt cache hits
40.9%
About $1.22 per 1M input tokens with caching
Request success (24h)
98.4%
Reliable

Cache hits are the share of input tokens recently served from a prompt cache across all requests. They are not a forecast of your own hit rate.

Hourly request success · last 72 hoursLowest hour 94.1%
  • Healthy (98% or more)
  • Degraded (90–98%)
  • Down (under 90%)
  • No requests

Developers

Use it through the API

Noumen’s API is OpenAI-compatible: any Chat Completions client works with a base URL and a key.

Read the quickstart
curl https://www.noumen.xyz/api/v1/chat/completions \
  -H "Authorization: Bearer $NOUMEN_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "google/gemini-4-argon",
    "reasoning_effort": "medium",
    "messages": [{ "role": "user", "content": "Explain this stack trace." }]
  }'
import os
from openai import OpenAI

client = OpenAI(base_url="https://www.noumen.xyz/api/v1", api_key=os.environ["NOUMEN_API_KEY"])
reply = client.chat.completions.create(
    model="google/gemini-4-argon",
    messages=[{"role": "user", "content": "Explain this stack trace."}],
)
print(reply.choices[0].message.content)

Google

Other Google models, starting with the same family.