GPT-5.4 vs GPT-6 Sol
Wins 1 of 8 areas
Following instructions
Wins 6 of 8 areas
Coding · Agents · Reasoning · Facts · Math · Long documents
GPT-6 Sol is the stronger all-rounder.GPT-5.4 is better at following instructions.
Scores updated · 35 tests both models report · How we compare
Where each one wins
Tests won in each of the eight areas we test. 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 thinking04GPT-6 Sol4 of 5 tests · 1 tie
- CodingWriting and fixing software03GPT-6 Sol3 of 4 tests · 1 tie
- FactsGetting facts right instead of making them up03GPT-6 Sol3 of 4 tests · 1 tie
- MathCompetition and research-level math03GPT-6 Sol3 of 3 tests
- AgentsCarrying out multi-step tasks on its own02GPT-6 Sol2 of 2 tests
- Long documentsFinding answers in very long texts01GPT-6 Sol1 of 1 test
- Following instructionsDoing exactly what it is asked10GPT-5.41 of 1 test
- Images and chartsUnderstanding pictures, charts and video11Even1 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.
GPT-6 Sol costs 31% less for the same work.
The biggest differences
The tests each model wins by the widest margin, up to three each. Scores are out of 100.
Where GPT-6 Sol pulls ahead
- Extremely hard research-level math problemsFrontierMath Tier 4+41points ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+31.6points ahead
- Abstract visual puzzles that people can solveARC-AGI-2+16.3points ahead
Where GPT-5.4 pulls ahead
- Follows detailed instructions exactlyLiveBench · Instruction Following+1.6points ahead
Every test, side by side
All 35 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingGPT-6 Sol
- LMArena · WebDevGPT-6 Sol by 211 rating points14621673+211 rating
- LiveBench · CodingGPT-6 Sol by 4.377.581.8+4.3
- SciCodeGPT-6 Sol by 156.657.6+1
- LiveBench · Agentic Codingtie53.852.9tie
AgentsGPT-6 Sol
- AA IT-Bench SREGPT-6 Sol by 14.934.549.4+14.9
- GDPValGPT-6 Sol by 13.137.350.4+13.1
ReasoningGPT-6 Sol
- ARC-AGI-2GPT-6 Sol by 16.373.389.6+16.3
- CritPtGPT-6 Sol by 7.523.430.9+7.5
- ARC-AGI-3GPT-6 Sol by 4.40.24.6+4.4
- Humanity's Last ExamGPT-6 Sol by 4.243.747.9+4.2
- LiveBench · Reasoningtie88.188.7tie
FactsGPT-6 Sol
- AA-Omniscience · Non-hallucinationGPT-6 Sol by 31.68.339.9+31.6
- SimpleQA VerifiedGPT-6 Sol by 15.645.160.7+15.6
- AA-Omniscience · AccuracyGPT-6 Sol by 3.650.954.5+3.6
- Vectara HHEM hallucination ratelower is bettertie76.5tie
Images and chartsEven
- LMArena · VisionGPT-5.4 by 68 rating points13011233+68 rating
- MMMU-ProGPT-6 Sol by 4.578.482.9+4.5
MathGPT-6 Sol
- FrontierMath Tier 4GPT-6 Sol by 414990+41
- FrontierMath Tiers 1-3 (v2)GPT-6 Sol by 11.278.689.8+11.2
- LiveBench · MathematicsGPT-6 Sol by 2.294.296.4+2.2
Following instructionsGPT-5.4
- LiveBench · Instruction FollowingGPT-5.4 by 1.670.268.6+1.6
Other results13 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.
- AA-OmniscienceGPT-6 Sol by 21.35.827.1+21.3
- DeepSWE 1.1GPT-6 Sol by 16.85268.8+16.8
- HealthBench Professional length-adjustedGPT-6 Sol by 12.748.160.8+12.7
- vectara_avg_summary_lengthGPT-5.4 by 10.381.771.4+10.3
- AA IntelligenceGPT-6 Sol by 8.63947.6+8.6
- livebench_languageGPT-6 Sol by 2.782.685.3+2.7
- livebench_data_analysisGPT-6 Sol by 1.979.381.2+1.9
- ARC-AGI-1GPT-6 Sol by 1.893.795.5+1.8
- HealthBench Hard length-adjustedGPT-6 Sol by 129.130.1+1
- HealthBench length-adjustedtie5453.2tie
- vectara_factual_consistencytie9393.5tie
- HealthBench Consensus length-adjustedtie96.396.2tie
- vectara_answer_ratetie99.9100tie
Questions people ask
Which is better, GPT-5.4 or GPT-6 Sol?
GPT-6 Sol wins six of the eight areas we test: coding, agents, reasoning, facts, math and long documents. GPT-5.4 wins following instructions. They are level on images and charts.
Which is better for coding?
GPT-6 Sol. It wins 3 of the 4 coding tests both models report; GPT-5.4 wins none, and 1 is a tie.
Which is cheaper?
GPT-5.4 costs $2.50 per million input tokens and $15.00 per million output tokens; GPT-6 Sol costs $2.00 and $10.00. That makes GPT-6 Sol about 31% cheaper for the same work.
How do you compare the two?
We use the 35 benchmark tests both models have published scores on. The verdict counts the 22 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 13 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.