Solar System
loads medium
Tech. quality74
UX84
Time4 min 23 s
Cost$0.30
Compare →
Gemini 4 Argon
Loads without errors5/5#1 of 26
Tech. quality91/100#17 of 26
UX88/100#8 of 26
Total cost$1.68
Average time5 min 05 s#17 of 26
Total tokens169,955
API price: $2 / $10 per 1M input / output tokens · as of 17/10/2026 source ↗
What is Gemini 4 Argon good for?
Google's flagship reasoning model – for complex software engineering, maths and science, multi-step agents and very long outputs.
Description based on the maker · source ↗
- Maker
- Released
- 30/09/2026
- Context window
- 1M tokens
- Model weights
- closed, the model is available through the API only
- Reasoning levels
- low medium high
How did Gemini 4 Argon do in the AI Lab tests?
- 5 of 5 outputs load without errors.
- Its best technical result was on the “New York & Central Park map” prompt (98/100).
- It implemented the chess rules flawlessly – passing the objective perft test on all 6 positions.
- Average time per prompt 5 min 05 s#17 of 26
- Average cost per prompt $0.34#19 of 26
- Technical quality 91/100#17 of 26
- Automatic UX rating 88/100#8 of 26
Outputs
All outputs of Gemini 4 Argon on the 5 prompts every other model got as well.
GTA-style game
loads medium
Tech. quality89
UX100
Time5 min 20 s
Cost$0.35
Compare →
Monitor test
loads medium
Tech. quality97
UX95
Time3 min 52 s
Cost$0.27
Compare →
New York & Central Park map
loads medium
Tech. quality98
UX84
Time6 min 34 s
Cost$0.40
Compare →
Chess with an engine
loads medium
Tech. quality96
UX76
Time5 min 17 s
Cost$0.37
Compare →
Cost = tokens priced with the API price list valid on the date shown next to the cost (prices change). ≈ = estimate (only the total token count is known), no ≈ = exact from input / output tokens. “Loads” = the page opens without errors; technical quality also checks behaviour during use – judge the quality in the viewer.