Gemini 4 Argon
Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents
- Strongest at
- Computer use Top 2% of the past year's models · 2 of 3 benchmarks
- Price per 1M tokens
- $2 / $10 input / output · Google
- Context window
- 1.05M tokens in · 65.5K out
Capability profile
Scored against 90 models released in the past year. 50 is the average model; further out is better.
Where these scores come from
Each capability combines the published benchmarks listed here, rescaled so 50 is the reference average and 15 points is one standard deviation. A hollow dot rests on a single benchmark.
- Coding FrontierCode, MirrorCode
- Writing EQ-Bench Creative Writing v3, EQ-Bench Longform Writing, Lech Mazur Writing comparison
- Reading & documents GDP.pdf (Surge AI)
- Reasoning ARC-AGI-2, Chess Puzzles, Mystery Game Puzzles
- Math OTIS Mock AIME 2024-2025, FrontierMath-Tiers-1-3-v2-Private, FrontierMath-Tier-4-v2-Private, ProofBench
- Science GPQA diamond
- Visual reasoning Blueprint-Bench 2, GDP.pdf (Surge AI)
- Computer use APEX-Agents, AutomationBench 1.0.6
Price against capability
The shaded corner holds anything better and cheaper. Prices blend input and output 3 to 1.
Nothing from the past year scores higher on Computer use for less.
3 models from the past year score higher on Coding for less: DeepSeek V4.1 Flash, Muse Spark 1.3, DeepSeek V3.2.
5 models from the past year score higher on Writing for less: Muse Spark 1.3, Gemini 3.8 Flash, GLM-5.3 and 2 more.
Nothing from the past year scores higher on Reading & documents for less.
Nothing from the past year scores higher on Reasoning for less.
Nothing from the past year scores higher on Math for less.
Nothing from the past year scores higher on Science for less.
Nothing from the past year scores higher on Visual reasoning for less.
Google $2 / $10 · OpenRouter $2 / $10 · Cached input $0.1
Epoch benchmark boards
Best result on each board it has run. The tick marks the leader.
10 more boards
Google's past year
38 release days in the past year, about one every 10 days. The next is expected around Oct 31, 2026. Forecasts ↗
More from the lab
Launch announcement · Oct 13, 2026Gemini 4 Argon: Built to solve your hardest problemsLaunch announcement · Oct 13, 2026Gemini 4 Argon is now available for developers.Related article · Oct 13, 2026Gemini 4 Argon comes to the Gemini app.Related article · Oct 13, 2026Gemini 4 Argon for RoboticsRelated article · Oct 23, 2026See what 5 builders are making with Gemini 4 ArgonSpecifications
Specifications and prices from Models.dev; availability and limits may differ by provider. Benchmarks from Epoch AI (CC BY 4.0), LiveBench and the other sources named above.