Gemini 4 Argon vs GPT-6.1 Sol
Wins 2 of 7 areas
Agents · Facts
Wins 4 of 7 areas
Reasoning · Math · Long documents · Following instructions
GPT-6.1 Sol wins more areas, narrowly.Gemini 4 Argon is better at agents.
Both rank among the ten best models we track in facts.
Scores updated · 24 tests both models report · How we compare
Where each one wins
Tests won in each of the seven areas where both have results. Each piece is one test, so a longer bar means more evidence; grey means the two scored within a point of each other.
- ReasoningHard problems that need careful thinking13GPT-6.1 Sol3 of 4 tests
- MathCompetition and research-level math02GPT-6.1 Sol2 of 2 tests
- Long documentsFinding answers in very long texts01GPT-6.1 Sol1 of 1 test
- Following instructionsDoing exactly what it is asked01GPT-6.1 Sol1 of 1 test
- AgentsCarrying out multi-step tasks on its own20Gemini 4 Argon2 of 2 tests
- FactsGetting facts right instead of making them up21Gemini 4 Argon2 of 3 tests
- CodingWriting and fixing software11Even1 each
Long documents and following instructions rest on a single test each.
What it costs
Prices per million tokens, roughly 750,000 words. The bars show the cost of a million tokens read plus a million written.
The biggest differences
The three tests each model wins by the widest margin. Scores are out of 100.
Where GPT-6.1 Sol pulls ahead
- Extremely hard research-level math problemsFrontierMath Tier 4+21.9points ahead
- Answers hard knowledge questions correctlyAA-Omniscience · Accuracy+12.2points ahead
- Unpublished advanced math problemsFrontierMath Tiers 1-3 (v2)+7.7points ahead
Where Gemini 4 Argon pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+39.2points ahead
- Code for real scientific research problemsSciCode+7.6points ahead
- Very hard expert questions across many subjectsHumanity's Last Exam+4.2points ahead
Every test, side by side
All 24 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingEven
- SciCodeGemini 4 Argon by 7.661.854.2+7.6
- LMArena · WebDevGPT-6.1 Sol by 72 rating points16791751+72 rating
AgentsGemini 4 Argon
- GDPValGemini 4 Argon by 1.556.254.7+1.5
- Terminal-Bench 4.0Gemini 4 Argon by 157.156.1+1
ReasoningGPT-6.1 Sol
- ARC-AGI-3GPT-6.1 Sol by 6.446.352.7+6.4
- CritPtGPT-6.1 Sol by 4.627.131.7+4.6
- Humanity's Last ExamGemini 4 Argon by 4.257.152.9+4.2
- ARC-AGI-2GPT-6.1 Sol by 2.991.394.2+2.9
FactsGemini 4 Argon
- AA-Omniscience · Non-hallucinationGemini 4 Argon by 39.284.945.7+39.2
- AA-Omniscience · AccuracyGPT-6.1 Sol by 12.249.962.1+12.2
- SimpleQA VerifiedGemini 4 Argon by 2.17673.9+2.1
MathGPT-6.1 Sol
- FrontierMath Tier 4GPT-6.1 Sol by 21.978100+21.9
- FrontierMath Tiers 1-3 (v2)GPT-6.1 Sol by 7.78693.7+7.7
Following instructionsGPT-6.1 Sol
- Multi-ChallengeGPT-6.1 Sol by 3.378.782+3.3
Other results9 tests, not counted
Tests outside the eight areas. They are not counted above: several are summary scores built from other tests, or the same test under another name.
- HealthBench HardGemini 4 Argon by 9.245.436.2+9.2
- DeepSWE 1.1Gemini 4 Argon by 2.777.975.2+2.7
- OSWorld 2.0GPT-6.1 Sol by 2.269.271.4+2.2
- ARC-AGI-1Gemini 4 Argon by 298.596.5+2
- AA-Omnisciencetie42.441.5tie
- AA Intelligencetie52.651.8tie
- Terminal-Bench 4.0tie57.458.2tie
- Terminal-Bench-Science 0.1tie57.657tie
- HiL-Bench (Tools-allowed)tie6262.3tie
Questions people ask
Which is better, Gemini 4 Argon or GPT-6.1 Sol?
GPT-6.1 Sol wins four of the seven areas where both have results: reasoning, math, long documents and following instructions. Gemini 4 Argon wins agents and facts. They are level on coding.
Which is better for coding?
Neither. They win 1 coding test each of the 2 both models report.
Which is cheaper?
Gemini 4 Argon costs $2.00 per million input tokens and $10.00 per million output tokens; GPT-6.1 Sol costs $2.00 and $10.00. For a million tokens read plus a million written, they cost about the same.
How do you compare the two?
We use the 24 benchmark tests both models have published scores on. The verdict counts the 15 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 9 are listed but not counted, because several are summary scores or repeat a test. Each score is the one shown on the model's own page.