GPT-5.5 vs GPT-6 Sol
Wins 1 of 8 areas
Following instructions
Wins 4 of 8 areas
Coding · Agents · Reasoning · Math
GPT-6 Sol wins more areas, narrowly.GPT-5.5 is better at following instructions.
Both rank among the ten best models we track in math.
Scores updated · 38 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 thinking15GPT-6 Sol5 of 6 tests
- AgentsCarrying out multi-step tasks on its own03GPT-6 Sol3 of 3 tests
- MathCompetition and research-level math02GPT-6 Sol2 of 3 tests · 1 tie
- CodingWriting and fixing software12GPT-6 Sol2 of 4 tests · 1 tie
- Following instructionsDoing exactly what it is asked10GPT-5.51 of 1 test
- FactsGetting facts right instead of making them up22Even2 each
- Images and chartsUnderstanding pictures, charts and video11Even1 each
- Long documentsFinding answers in very long texts00Even0 each · 1 tie
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 66% less for the same work.
The biggest differences
The three tests each model wins by the widest margin. Scores are out of 100.
Where GPT-6 Sol pulls ahead
- Complex command-line tasks across many fieldsTerminal-Bench 4.0+29.3points ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+28.9points ahead
- Extremely hard research-level math problemsFrontierMath Tier 4+17.5points ahead
Where GPT-5.5 pulls ahead
- Answers hard knowledge questions correctlyAA-Omniscience · Accuracy+3.5points ahead
- Short factual questions, answered correctlySimpleQA Verified+2.3points ahead
- Follows detailed instructions exactlyLiveBench · Instruction Following+2.1points ahead
Every test, side by side
All 38 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 186 rating points14871673+186 rating
- SciCodeGPT-6 Sol by 1.855.857.6+1.8
- LiveBench · Agentic CodingGPT-5.5 by 1.15452.9+1.1
- LiveBench · Codingtie82.281.8tie
AgentsGPT-6 Sol
- Terminal-Bench 4.0GPT-6 Sol by 29.314.643.9+29.3
- GDPValGPT-6 Sol by 7.942.550.4+7.9
- AA IT-Bench SREGPT-6 Sol by 3.645.849.4+3.6
ReasoningGPT-6 Sol
- ARC-AGI-2GPT-6 Sol by 4.68589.6+4.6
- ARC-AGI-3GPT-6 Sol by 4.20.44.6+4.2
- SimpleBenchGPT-6 Sol by 4.16973.1+4.1
- CritPtGPT-6 Sol by 3.827.130.9+3.8
- Humanity's Last ExamGPT-6 Sol by 2.145.847.9+2.1
- LiveBench · ReasoningGPT-5.5 by 189.788.7+1
FactsEven
- AA-Omniscience · Non-hallucinationGPT-6 Sol by 28.91139.9+28.9
- AA-Omniscience · AccuracyGPT-5.5 by 3.55854.5+3.5
- Vectara HHEM hallucination ratelower is betterGPT-6 Sol by 2.89.36.5+2.8
- SimpleQA VerifiedGPT-5.5 by 2.36360.7+2.3
Images and chartsEven
- LMArena · VisionGPT-5.5 by 62 rating points12951233+62 rating
- MMMU-ProGPT-6 Sol by 379.982.9+3
MathGPT-6 Sol
- FrontierMath Tier 4GPT-6 Sol by 17.572.590+17.5
- FrontierMath Tiers 1-3 (v2)GPT-6 Sol by 4.585.389.8+4.5
- LiveBench · Mathematicstie95.996.4tie
Following instructionsGPT-5.5
- LiveBench · Instruction FollowingGPT-5.5 by 2.170.768.6+2.1
Other results14 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.
- Agents' Last ExamGPT-6 Sol by 29.826.656.4+29.8
- AA IntelligenceGPT-6 Sol by 9.238.447.6+9.2
- HealthBench Professional length-adjustedGPT-6 Sol by 951.860.8+9
- AA-OmniscienceGPT-6 Sol by 6.620.527.1+6.6
- vectara_avg_summary_lengthGPT-5.5 by 58.2 rating points129.671.4+58.2 rating
- HealthBench length-adjustedGPT-5.5 by 3.356.553.2+3.3
- vectara_factual_consistencyGPT-6 Sol by 2.890.793.5+2.8
- livebench_languageGPT-5.5 by 2.187.485.3+2.1
- DeepSWE 1.1GPT-6 Sol by 1.86768.8+1.8
- HealthBench Hard length-adjustedGPT-5.5 by 1.431.530.1+1.4
- HealthBench Consensus length-adjustedtie95.696.2tie
- ARC-AGI-1tie9595.5tie
- livebench_data_analysistie81.681.2tie
- vectara_answer_ratetie100100tie
Questions people ask
Which is better, GPT-5.5 or GPT-6 Sol?
GPT-6 Sol wins four of the eight areas we test: coding, agents, reasoning and math. GPT-5.5 wins following instructions. They are level on facts, images and charts and long documents.
Which is better for coding?
GPT-6 Sol. It wins 2 of the 4 coding tests both models report; GPT-5.5 wins 1, and 1 is a tie.
Which is cheaper?
GPT-5.5 costs $5.00 per million input tokens and $30.00 per million output tokens; GPT-6 Sol costs $2.00 and $10.00. That makes GPT-6 Sol about 66% cheaper for the same work.
How do you compare the two?
We use the 38 benchmark tests both models have published scores on. The verdict counts the 24 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 14 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.