Gemini 3.7 Flash vs GPT-5.6 Sol
Wins 2 of 8 areas
Images and charts · Following instructions
Wins 5 of 8 areas
Coding · Agents · Reasoning · Math · Long documents
GPT-5.6 Sol is the stronger all-rounder.Gemini 3.7 Flash is cheaper and better at images and charts.
Scores updated · 44 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-5.6 Sol4 of 5 tests · 1 tie
- AgentsCarrying out multi-step tasks on its own03GPT-5.6 Sol3 of 3 tests
- MathCompetition and research-level math03GPT-5.6 Sol3 of 3 tests
- CodingWriting and fixing software13GPT-5.6 Sol3 of 5 tests · 1 tie
- Long documentsFinding answers in very long texts01GPT-5.6 Sol1 of 1 test
- Images and chartsUnderstanding pictures, charts and video30Gemini 3.7 Flash3 of 3 tests
- Following instructionsDoing exactly what it is asked10Gemini 3.7 Flash1 of 1 test
- FactsGetting facts right instead of making them up11Even1 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.
Gemini 3.7 Flash costs 81% 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-5.6 Sol pulls ahead
- Extremely hard research-level math problemsFrontierMath Tier 4+46.4points ahead
- Complex command-line tasks across many fieldsTerminal-Bench 4.0+26.3points ahead
- Research-level physics problemsCritPt+18points ahead
Where Gemini 3.7 Flash pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+27.7points ahead
- Follows detailed instructions exactlyLiveBench · Instruction Following+8points ahead
- Reasoning about charts from research papersCharXiv (RQ)+2.9points ahead
Every test, side by side
All 44 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingGPT-5.6 Sol
- LiveBench · CodingGPT-5.6 Sol by 578.983.9+5
- LMArena · WebDevGPT-5.6 Sol by 26 rating points15921618+26 rating
- Terminal-Bench 2.1GPT-5.6 Sol by 2.285.888+2.2
- LiveBench · Agentic CodingGemini 3.7 Flash by 2.158.356.2+2.1
- SciCodetie57.257.1tie
AgentsGPT-5.6 Sol
- Terminal-Bench 4.0GPT-5.6 Sol by 26.313.639.9+26.3
- τ-Bench V3 · BankingGPT-5.6 Sol by 11.532.844.3+11.5
- GDPValGPT-5.6 Sol by 1144.655.6+11
ReasoningGPT-5.6 Sol
- CritPtGPT-5.6 Sol by 1814.332.3+18
- ARC-AGI-2GPT-5.6 Sol by 7.984.692.5+7.9
- LiveBench · ReasoningGPT-5.6 Sol by 3.987.891.7+3.9
- Humanity's Last ExamGPT-5.6 Sol by 1.647.949.5+1.6
- GPQA Diamondtie94.594.1tie
FactsEven
- AA-Omniscience · Non-hallucinationGemini 3.7 Flash by 27.735.57.8+27.7
- AA-Omniscience · AccuracyGPT-5.6 Sol by 4.155.359.4+4.1
- SimpleQA Verifiedtie69.269.7tie
Images and chartsGemini 3.7 Flash
- LMArena · VisionGemini 3.7 Flash by 36 rating points13151279+36 rating
- CharXiv (RQ)Gemini 3.7 Flash by 2.988.785.8+2.9
- MMMU-ProGemini 3.7 Flash by 2.185.583.4+2.1
MathGPT-5.6 Sol
- FrontierMath Tier 4GPT-5.6 Sol by 46.436.683+46.4
- FrontierMath Tiers 1-3 (v2)GPT-5.6 Sol by 17.571.689.1+17.5
- LiveBench · MathematicsGPT-5.6 Sol by 2.793.596.2+2.7
Following instructionsGemini 3.7 Flash
- LiveBench · Instruction FollowingGemini 3.7 Flash by 879.971.8+8
Other results20 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-5.6 Sol by 26.426.352.7+26.4
- Terminal-Bench 4.0GPT-5.6 Sol by 26.111.237.3+26.1
- GDPval-AA v2 EloGPT-5.6 Sol by 228 rating points14821710+228 rating
- Terminal-bench 3.0GPT-5.6 Sol by 19.514.934.4+19.5
- GDPval-AA v2GPT-5.6 Sol by 186 rating points15251711+186 rating
- OSWorld 2.0GPT-5.6 Sol by 14.747.962.6+14.7
- AA Agentic IndexGPT-5.6 Sol by 14.136.450.5+14.1
- OSWorld 2.0 (partial)GPT-5.6 Sol by 1250.662.6+12
- livebench_data_analysisGPT-5.6 Sol by 11.86879.8+11.8
- AA IntelligenceGPT-5.6 Sol by 7.939.147+7.9
- DeepSWE 1.1GPT-5.6 Sol by 7.765.373+7.7
- GDP.PDF (All pass rate)GPT-5.6 Sol by 63440+6
- AA-OmniscienceGemini 3.7 Flash by 4.526.522+4.5
- GDP (Surge AI)Gemini 3.7 Flash by 3.33430.7+3.3
- LVBenchGemini 3.7 Flash by 3.385.482.1+3.3
- Artificial AnalysisGPT-5.6 Sol by 35659+3
- livebench_languageGPT-5.6 Sol by 2.285.587.7+2.2
- Artificial Analysis Coding IndexGPT-5.6 Sol by 1.376.177.4+1.3
- ARC-AGI-1GPT-5.6 Sol by 195.596.5+1
- HLE-Verifiedtie53.654.5tie
Questions people ask
Which is better, Gemini 3.7 Flash or GPT-5.6 Sol?
GPT-5.6 Sol wins five of the eight areas we test: coding, agents, reasoning, math and long documents. Gemini 3.7 Flash wins images and charts and following instructions, and costs 81% less. They are level on facts.
Which is better for coding?
GPT-5.6 Sol. It wins 3 of the 5 coding tests both models report; Gemini 3.7 Flash wins 1, and 1 is a tie.
Which is cheaper?
Gemini 3.7 Flash costs $0.75 per million input tokens and $3.75 per million output tokens; GPT-5.6 Sol costs $4.00 and $20.00. That makes Gemini 3.7 Flash about 81% cheaper for the same work.
How do you compare the two?
We use the 44 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 20 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.