Qwen3.6 27B vs Qwen3.7 Plus Preview
Wins 1 of 8 areas
Long documents
Wins 6 of 8 areas
Coding · Reasoning · Facts · Images and charts · Math · Following instructions
Qwen3.7 Plus Preview is the stronger all-rounder.Qwen3.6 27B is better at long documents.
Scores updated · 43 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.
- CodingWriting and fixing software05Qwen3.7 Plus Preview5 of 7 tests · 2 ties
- ReasoningHard problems that need careful thinking03Qwen3.7 Plus Preview3 of 3 tests
- MathCompetition and research-level math02Qwen3.7 Plus Preview2 of 3 tests · 1 tie
- FactsGetting facts right instead of making them up02Qwen3.7 Plus Preview2 of 2 tests
- Images and chartsUnderstanding pictures, charts and video12Qwen3.7 Plus Preview2 of 3 tests
- Following instructionsDoing exactly what it is asked01Qwen3.7 Plus Preview1 of 1 test
- Long documentsFinding answers in very long texts10Qwen3.6 27B1 of 1 test
- AgentsCarrying out multi-step tasks on its own11Even1 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.
Qwen3.7 Plus Preview costs 52% less for the same work.
The biggest differences
The three tests each model wins by the widest margin. Scores are out of 100.
Where Qwen3.7 Plus Preview pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+21.6points ahead
- Very hard expert questions across many subjectsHumanity's Last Exam+12.5points ahead
- Hard command-line tasks in a real terminalTerminal-Bench Hard+12.2points ahead
Where Qwen3.6 27B pulls ahead
- Real work tasks from 44 professionsGDPVal+10.9points ahead
- Reasons across sets of long documentsAA-LCR+4.3points ahead
- Reads and reasons about text in imagesOCRBenchv2+1.2points ahead
Every test, side by side
All 43 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingQwen3.7 Plus Preview
- Terminal-Bench HardQwen3.7 Plus Preview by 12.234.847+12.2
- LiveCodeBench v6Qwen3.7 Plus Preview by 5.783.989.6+5.7
- SWE-bench MultilingualQwen3.7 Plus Preview by 4.571.375.8+4.5
- SWE-bench ProQwen3.7 Plus Preview by 4.153.557.6+4.1
- SciCodeQwen3.7 Plus Preview by 3.342.846.1+3.3
- SWE-bench Verifiedtie77.277.7tie
- Terminal-Bench 2.1tie60.761tie
AgentsEven
- GDPValQwen3.6 27B by 10.924.413.5+10.9
- Terminal-Bench 4.0Qwen3.7 Plus Preview by 101+1
- τ-Bench V3 · Bankingtie16.717.5tie
ReasoningQwen3.7 Plus Preview
- Humanity's Last ExamQwen3.7 Plus Preview by 12.523.135.6+12.5
- CritPtQwen3.7 Plus Preview by 81.19.1+8
- GPQA DiamondQwen3.7 Plus Preview by 5.884.290+5.8
FactsQwen3.7 Plus Preview
- AA-Omniscience · Non-hallucinationQwen3.7 Plus Preview by 21.650.772.3+21.6
- AA-Omniscience · AccuracyQwen3.7 Plus Preview by 2.919.622.5+2.9
Images and chartsQwen3.7 Plus Preview
- CharXiv (RQ)Qwen3.7 Plus Preview by 7.578.485.9+7.5
- MMMU-ProQwen3.7 Plus Preview by 5.974.680.5+5.9
- OCRBenchv2Qwen3.6 27B by 1.268.367.1+1.2
MathQwen3.7 Plus Preview
- HMMT Feb. 2026Qwen3.7 Plus Preview by 8.684.392.9+8.6
- IMOAnswerBenchQwen3.7 Plus Preview by 5.280.886+5.2
- FrontierMath Tiers 1-3 (v2)tie35.134.4tie
Following instructionsQwen3.7 Plus Preview
- IFBenchQwen3.7 Plus Preview by 10.467.678+10.4
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.
- SimpleVQAQwen3.7 Plus Preview by 25.656.181.7+25.6
- AA-OmniscienceQwen3.7 Plus Preview by 21.1-201.1+21.1
- Terminal-Bench 2.0Qwen3.7 Plus Preview by 1159.370.3+11
- AndroidWorldQwen3.7 Plus Preview by 10.770.381+10.7
- QwenClawBenchQwen3.7 Plus Preview by 8.453.461.8+8.4
- ERQAQwen3.7 Plus Preview by 7.362.569.8+7.3
- JobBenchQwen3.7 Plus Preview by 5.821.827.6+5.8
- SuperGPQAQwen3.7 Plus Preview by 5.46671.4+5.4
- NL2RepoQwen3.7 Plus Preview by 4.936.241.1+4.9
- AA IntelligenceQwen3.7 Plus Preview by 3.821.425.2+3.8
- DeepSWE 1.1Qwen3.7 Plus Preview by 3.213.316.5+3.2
- RealWorldQAQwen3.7 Plus Preview by 2.884.186.9+2.8
- MMLU-ProQwen3.7 Plus Preview by 2.386.288.5+2.3
- Artificial Analysis Coding IndexQwen3.7 Plus Preview by 2.253.755.9+2.2
- Claw Eval (pass@3)Qwen3.7 Plus Preview by 2.160.662.7+2.1
- τ²-Bench Telecom (AA run)Qwen3.6 27B by 1.294.293+1.2
- MMLU-ReduxQwen3.7 Plus Preview by 193.594.5+1
- VideoMMMUQwen3.7 Plus Preview by 184.485.4+1
- MLVUtie86.687.4tie
- AA Agentic Indextie20.119.7tie
Questions people ask
Which is better, Qwen3.6 27B or Qwen3.7 Plus Preview?
Qwen3.7 Plus Preview wins six of the eight areas we test: coding, reasoning, facts, images and charts, math and following instructions. Qwen3.6 27B wins long documents. They are level on agents.
Which is better for coding?
Qwen3.7 Plus Preview. It wins 5 of the 7 coding tests both models report; Qwen3.6 27B wins none, and 2 are ties.
Which is cheaper?
Qwen3.6 27B costs $0.60 per million input tokens and $3.60 per million output tokens; Qwen3.7 Plus Preview costs $0.40 and $1.60. That makes Qwen3.7 Plus Preview about 52% cheaper for the same work.
How do you compare the two?
We use the 43 benchmark tests both models have published scores on. The verdict counts the 23 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.