GPT-5 vs Qwen3.6 27B
Wins 3 of 8 areas
Reasoning · Images and charts · Following instructions
Wins 1 of 8 areas
Coding
GPT-5 wins more areas, narrowly.Qwen3.6 27B is cheaper and better at coding.
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 thinking30GPT-53 of 3 tests
- Images and chartsUnderstanding pictures, charts and video21GPT-52 of 4 tests · 1 tie
- Following instructionsDoing exactly what it is asked10GPT-51 of 1 test
- CodingWriting and fixing software15Qwen3.6 27B5 of 7 tests · 1 tie
- AgentsCarrying out multi-step tasks on its own11Even1 each
- FactsGetting facts right instead of making them up11Even1 each
- MathCompetition and research-level math11Even1 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.
Qwen3.6 27B costs 63% 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 pulls ahead
- Answers hard knowledge questions correctlyAA-Omniscience · Accuracy+20.7points ahead
- Unpublished advanced math problemsFrontierMath Tiers 1-3 (v2)+20.3points ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+5.5points ahead
Where Qwen3.6 27B pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+32.9points ahead
- Command-line tasks in a real terminalTerminal-Bench 2.1+25.5points ahead
- Fixes real GitHub issues in many programming languagesSWE-bench Multilingual+16points 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.
CodingQwen3.6 27B
- Terminal-Bench 2.1Qwen3.6 27B by 25.535.260.7+25.5
- SWE-bench MultilingualQwen3.6 27B by 1655.371.3+16
- SWE-bench ProQwen3.6 27B by 11.741.853.5+11.7
- LiveCodeBench v6GPT-5 by 3.18783.9+3.1
- SWE-bench VerifiedQwen3.6 27B by 2.374.977.2+2.3
- Terminal-Bench HardQwen3.6 27B by 2.232.634.8+2.2
- SciCodetie42.942.8tie
AgentsEven
- τ-Bench V3 · BankingGPT-5 by 5.422.116.7+5.4
- GDPValQwen3.6 27B by 3.42124.4+3.4
ReasoningGPT-5
- Humanity's Last ExamGPT-5 by 5.428.523.1+5.4
- CritPtGPT-5 by 4.65.71.1+4.6
- GPQA DiamondGPT-5 by 1.285.484.2+1.2
FactsEven
- AA-Omniscience · Non-hallucinationQwen3.6 27B by 32.917.850.7+32.9
- AA-Omniscience · AccuracyGPT-5 by 20.740.319.6+20.7
Images and chartsGPT-5
- OCRBenchv2Qwen3.6 27B by 12.855.568.3+12.8
- CharXiv (RQ)GPT-5 by 2.781.178.4+2.7
- MMMUGPT-5 by 1.384.282.9+1.3
- MMMU-Protie74.274.6tie
MathEven
- FrontierMath Tiers 1-3 (v2)GPT-5 by 20.355.435.1+20.3
- IMOAnswerBenchQwen3.6 27B by 4.87680.8+4.8
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.
- Terminal-Bench 2.0Qwen3.6 27B by 24.135.259.3+24.1
- Artificial Analysis Coding IndexQwen3.6 27B by 15.937.853.7+15.9
- AA-OmniscienceGPT-5 by 11.3-8.7-20+11.3
- τ²-Bench Telecom (AA run)Qwen3.6 27B by 9.484.894.2+9.4
- AA Agentic IndexGPT-5 by 6.426.520.1+6.4
- HMMT Feb. 2025Qwen3.6 27B by 5.588.393.8+5.5
- ERQAGPT-5 by 3.265.762.5+3.2
- HMMT 2025GPT-5 by 2.693.390.7+2.6
- MMLU-ReduxGPT-5 by 1.895.393.5+1.8
- AA IntelligenceGPT-5 by 1.62321.4+1.6
- HMMT Nov. 2025Qwen3.6 27B by 1.589.290.7+1.5
- MMLU-Protie87.186.2tie
- VideoMMMUtie84.684.4tie
Questions people ask
Which is better, GPT-5 or Qwen3.6 27B?
GPT-5 wins three of the eight areas we test: reasoning, images and charts and following instructions. Qwen3.6 27B wins coding, and costs 63% less. They are level on agents, facts, math and long documents.
Which is better for coding?
Qwen3.6 27B. It wins 5 of the 7 coding tests both models report; GPT-5 wins 1, and 1 is a tie.
Which is cheaper?
GPT-5 costs $1.25 per million input tokens and $10.00 per million output tokens; Qwen3.6 27B costs $0.60 and $3.60. That makes Qwen3.6 27B about 63% 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.