GPT-5.5 vs Qwen3.7 Max
Wins 6 of 7 areas
Coding · Agents · Reasoning · Facts · Math · Long documents
Wins 1 of 7 areas
Following instructions
GPT-5.5 is the stronger all-rounder.Qwen3.7 Max is cheaper and better at following instructions.
Scores updated · 48 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.
- MathCompetition and research-level math50GPT-5.55 of 5 tests
- CodingWriting and fixing software51GPT-5.55 of 7 tests · 1 tie
- AgentsCarrying out multi-step tasks on its own41GPT-5.54 of 5 tests
- ReasoningHard problems that need careful thinking41GPT-5.54 of 5 tests
- FactsGetting facts right instead of making them up21GPT-5.52 of 3 tests
- Long documentsFinding answers in very long texts10GPT-5.51 of 1 test
- Following instructionsDoing exactly what it is asked02Qwen3.7 Max2 of 2 tests
Long documents rests on a single test.
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.
Qwen3.7 Max costs 71% 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-5.5 pulls ahead
- Extremely hard research-level math problemsFrontierMath Tier 4+38.3points ahead
- Customer-service tasks in a simulated bankτ-Bench V3 · Banking+27.2points ahead
- Answers hard knowledge questions correctlyAA-Omniscience · Accuracy+26.9points ahead
Where Qwen3.7 Max pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+63.4points ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+4.6points ahead
- Follows detailed instructions exactlyLiveBench · Instruction Following+3.3points ahead
Every test, side by side
All 48 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingGPT-5.5
- LiveBench · Agentic CodingGPT-5.5 by 10.45443.6+10.4
- Terminal-Bench 2.1GPT-5.5 by 9.884.374.5+9.8
- Terminal-Bench HardGPT-5.5 by 9.860.650.8+9.8
- LiveBench · CodingGPT-5.5 by 882.274.2+8
- SciCodeGPT-5.5 by 6.355.849.5+6.3
- SWE-bench ProQwen3.7 Max by 258.660.6+2
- LMArena · WebDevtie15131515tie
AgentsGPT-5.5
- τ-Bench V3 · BankingGPT-5.5 by 27.23911.8+27.2
- Terminal-Bench 4.0GPT-5.5 by 13.114.61.5+13.1
- GDPValGPT-5.5 by 11.142.731.6+11.1
- AA IT-Bench SREGPT-5.5 by 3.345.842.5+3.3
- MCP AtlasQwen3.7 Max by 1.175.376.4+1.1
ReasoningGPT-5.5
- CritPtGPT-5.5 by 13.727.113.4+13.7
- LiveBench · ReasoningGPT-5.5 by 6.489.783.3+6.4
- Humanity's Last ExamGPT-5.5 by 5.345.840.5+5.3
- SimpleBenchQwen3.7 Max by 1.46970.4+1.4
- GPQA DiamondGPT-5.5 by 1.293.592.3+1.2
FactsGPT-5.5
- AA-Omniscience · Non-hallucinationQwen3.7 Max by 63.41174.4+63.4
- AA-Omniscience · AccuracyGPT-5.5 by 26.95831.1+26.9
- SimpleQA VerifiedGPT-5.5 by 7.26355.8+7.2
MathGPT-5.5
- FrontierMath Tier 4GPT-5.5 by 38.372.534.1+38.3
- FrontierMath Tiers 1-3 (v2)GPT-5.5 by 20.785.364.6+20.7
- LiveBench · MathematicsGPT-5.5 by 10.695.985.3+10.6
- AIME 2026GPT-5.5 by 310097+3
- HMMT Feb. 2026GPT-5.5 by 1.498.597.1+1.4
Following instructionsQwen3.7 Max
- IFBenchQwen3.7 Max by 4.675.980.5+4.6
- LiveBench · Instruction FollowingQwen3.7 Max by 3.370.774+3.3
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.
- DeepSWEGPT-5.5 by 527018+52
- DeepSWE 1.1GPT-5.5 by 45.46721.6+45.4
- Toolathlon VerifiedGPT-5.5 by 23.873.549.7+23.8
- AA Agentic IndexGPT-5.5 by 13.437.323.9+13.4
- Terminal-Bench 2.0GPT-5.5 by 1382.769.7+13
- livebench_data_analysisGPT-5.5 by 9.881.671.8+9.8
- AA IntelligenceGPT-5.5 by 8.938.429.5+8.9
- Artificial Analysis Coding IndexGPT-5.5 by 8.974.966+8.9
- livebench_languageGPT-5.5 by 7.787.479.7+7.7
- AA-OmniscienceGPT-5.5 by 720.513.5+7
- JobBenchGPT-5.5 by 738.331.3+7
- LiveBenchGPT-5.5 by 6.480.774.3+6.4
- MLS Bench Litelower is betterQwen3.7 Max by 3.835.531.7+3.8
- NL2RepoGPT-5.5 by 3.550.747.2+3.5
- Finance Agent v2GPT-5.5 by 3.451.848.4+3.4
- HealthBenchGPT-5.5 by 256.554.5+2
- HMMT Nov. 2025GPT-5.5 by 1.596.595+1.5
- HLE (with tools)Qwen3.7 Max by 1.352.253.5+1.3
- SuperGPQAtie72.773.6tie
- τ²-Bench Telecom (AA run)tie93.994.7tie
Questions people ask
Which is better, GPT-5.5 or Qwen3.7 Max?
GPT-5.5 wins six of the seven areas where both have results: coding, agents, reasoning, facts, math and long documents. Qwen3.7 Max wins following instructions, and costs 71% less.
Which is better for coding?
GPT-5.5. It wins 5 of the 7 coding tests both models report; Qwen3.7 Max 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; Qwen3.7 Max costs $2.50 and $7.50. That makes Qwen3.7 Max about 71% cheaper for the same work.
How do you compare the two?
We use the 48 benchmark tests both models have published scores on. The verdict counts the 28 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.