Skip to content

Gemini 4 Argon vs GPT-6.1 Sol: Is OpenAI Falling Behind?

Google has made Gemini 4 Argon broadly available, and it edges past GPT-6.1 Sol on most benchmarks, including the Artificial Analysis Intelligence Index, 53 to 52. GPT-6.1 Sol still wins on cost: it runs the Index for less than half the price and keeps $2 and $10 pricing after Argon's introductory price ends on December 1. Here is how the two compare.

Benchmarks

BenchmarkGemini 4 ArgonGPT-6.1 SolDifference
DeepSWE v1.177.9%75.2%Argon +2.7 points
AutomationBench51.3%36.1%Argon +15.2 points
OSWorld 2.0 (offline, partial credit)69.2%71.4%Sol +2.2 points
Terminal-Bench Science 0.157.6%57.0%Argon +0.6 points

Artificial Analysis benchmarks

BenchmarkGemini 4 Argon (High)GPT-6.1 Sol (Max)Difference
Intelligence Index5352Argon +1
AA-Briefcase v1.11488 Elo1561 EloSol +73 Elo
GDPval-AA v2.11624 Elo1592 EloArgon +32 Elo
AutomationBench-AA78%65%Argon +13 points
Terminal-Bench 4.057%56%Argon +1 point
SciCode62%54%Argon +8 points
Humanity's Last Exam57%53%Argon +4 points
GDP.pdf22%31%Sol +9 points
CritPt27%32%Sol +5 points
AA-Omniscience4242Tie
AA-LCR v1.180%83%Sol +3 points

Cost and efficiency

MetricGemini 4 Argon (High)GPT-6.1 Sol (Max)Difference
Intelligence Index5352Argon +1
Cost per Intelligence Index task$1.99$0.72Sol about 64% cheaper
Cost to run the Intelligence Index$2,407$1,082Sol about 55% cheaper
Output tokens per task62K38KSol uses about 39% fewer
Reasoning tokens per task36K25KSol uses about 31% fewer
Output tokens across the Index113M67MSol uses about 41% fewer
Input, per 1M tokens$2$2Same for now
Cached input, per 1M tokens$0.10$0.10Same for now
Output, per 1M tokens$10$10Same for now

Pricing and specs

SpecificationGemini 4 ArgonGPT-6.1 SolDifference
Input now, per 1M tokens$2Introductory$2Same
Cached input now, per 1M tokens$0.10$0.10Same
Output now, per 1M tokens$10Introductory$10Same
Input from December 1, 2026$4$2Sol 50% cheaper
Output from December 1, 2026$20$10Sol 50% cheaper
Context windowAbout 1M tokens1.05M tokensNearly identical
Max outputUp to 1M tokens128K tokensArgon much larger
Thinking levelsLow, Medium, HighLow, Medium, High, Xhigh, MaxSol more granular
Text inputYesYesSame
Image inputYesYesSame
Audio and video inputYesNoArgon broader
APIGemini APIOpenAI APIBoth available