GPT-5.2 vs Qwen3.8 27B
Wins 3 of 8 areas
Reasoning · Math · Long documents
Wins 3 of 8 areas
Coding · Images and charts · Following instructions
The two are evenly matched.GPT-5.2 is better at reasoning and long documents; Qwen3.8 27B at images and charts and coding.
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.
- Images and chartsUnderstanding pictures, charts and video04Qwen3.8 27B4 of 4 tests
- CodingWriting and fixing software24Qwen3.8 27B4 of 7 tests · 1 tie
- Following instructionsDoing exactly what it is asked02Qwen3.8 27B2 of 2 tests
- ReasoningHard problems that need careful thinking41GPT-5.24 of 6 tests · 1 tie
- Long documentsFinding answers in very long texts10GPT-5.21 of 2 tests · 1 tie
- MathCompetition and research-level math10GPT-5.21 of 1 test
- AgentsCarrying out multi-step tasks on its own22Even2 each
- FactsGetting facts right instead of making them up11Even1 each
Math 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.
Qwen3.8 27B costs 78% less for the same work.
The biggest differences
The three tests each model wins by the widest margin. Scores are out of 100.
Where Qwen3.8 27B pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+50.9points ahead
- Customer-service tasks in a simulated bankτ-Bench V3 · Banking+36.9points ahead
- Long, multi-file coding tasks in real codebasesSWE-bench Pro+31.8points ahead
Where GPT-5.2 pulls ahead
- Answers hard knowledge questions correctlyAA-Omniscience · Accuracy+28.7points ahead
- Abstract visual puzzles that people can solveARC-AGI-2+10.5points ahead
- Fresh competition math problemsLiveBench · Mathematics+7points 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.
CodingQwen3.8 27B
- SWE-bench ProQwen3.8 27B by 31.829.961.7+31.8
- LMArena · WebDevQwen3.8 27B by 171 rating points14191590+171 rating
- LiveBench · Agentic CodingQwen3.8 27B by 11.150.361.4+11.1
- SWE-bench MultilingualQwen3.8 27B by 7.166.773.8+7.1
- SciCodeGPT-5.2 by 5.552.146.6+5.5
- SWE-bench VerifiedGPT-5.2 by 2.28077.8+2.2
- LiveBench · Codingtie76.175.7tie
AgentsEven
- τ-Bench V3 · BankingQwen3.8 27B by 36.911.148+36.9
- MCP AtlasQwen3.8 27B by 16.760.677.3+16.7
- BrowseCompGPT-5.2 by 4.265.861.6+4.2
- GDPValGPT-5.2 by 2.348.346+2.3
ReasoningGPT-5.2
- SimpleBenchQwen3.8 27B by 14.445.860.2+14.4
- ARC-AGI-2GPT-5.2 by 10.552.942.4+10.5
- CritPtGPT-5.2 by 6.211.65.4+6.2
- Humanity's Last ExamGPT-5.2 by 3.837.733.9+3.8
- LiveBench · ReasoningGPT-5.2 by 3.283.280+3.2
- GPQA Diamondtie90.390.5tie
FactsEven
- AA-Omniscience · Non-hallucinationQwen3.8 27B by 50.918.869.7+50.9
- AA-Omniscience · AccuracyGPT-5.2 by 28.744.315.6+28.7
Images and chartsQwen3.8 27B
- OCRBenchv2Qwen3.8 27B by 1350.563.5+13
- CharXiv (RQ)Qwen3.8 27B by 8.182.190.2+8.1
- LMArena · VisionQwen3.8 27B by 25 rating points12461271+25 rating
- MMMU-ProQwen3.8 27B by 1.774.676.3+1.7
Long documentsGPT-5.2
- LongBench v2GPT-5.2 by 2.254.552.3+2.2
- AA-LCRtie82.782tie
Following instructionsQwen3.8 27B
- LiveBench · Instruction FollowingQwen3.8 27B by 10.961.872.7+10.9
- IFBenchQwen3.8 27B by 4.175.479.5+4.1
Other results10 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.
- Artificial Analysis Coding IndexQwen3.8 27B by 19.448.768.1+19.4
- AA Agentic IndexGPT-5.2 by 13.760.246.5+13.7
- MATH-VisionQwen3.8 27B by 11.68394.6+11.6
- AA-OmniscienceGPT-5.2 by 9.1-0.9-10+9.1
- DeepSearchQA (F1)Qwen3.8 27B by 8.471.379.7+8.4
- livebench_languageGPT-5.2 by 5.479.874.3+5.4
- OmniDocBench 1.5Qwen3.8 27B by 5.485.791.1+5.4
- AA IntelligenceQwen3.8 27B by 3.330.433.7+3.3
- livebench_data_analysisGPT-5.2 by 1.678.276.6+1.6
- ARC-AGI-1Qwen3.8 27B by 1.386.287.5+1.3
Questions people ask
Which is better, GPT-5.2 or Qwen3.8 27B?
GPT-5.2 and Qwen3.8 27B each win three of the eight areas we test. GPT-5.2 wins reasoning, math and long documents; Qwen3.8 27B wins coding, images and charts and following instructions. Qwen3.8 27B costs 78% less. They are level on agents and facts.
Which is better for coding?
Qwen3.8 27B. It wins 4 of the 7 coding tests both models report; GPT-5.2 wins 2, and 1 is a tie.
Which is cheaper?
GPT-5.2 costs $1.75 per million input tokens and $14.00 per million output tokens; Qwen3.8 27B costs $0.50 and $3.00. That makes Qwen3.8 27B about 78% 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 28 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 10 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.