Kimi K2.5 vs Qwen3.7 Plus Preview
Wins 1 of 8 areas
Long documents
Wins 6 of 8 areas
Coding · Agents · Reasoning · Images and charts · Math · Following instructions
Qwen3.7 Plus Preview is the stronger all-rounder.Kimi K2.5 is better at long documents.
Scores updated · 41 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 software15Qwen3.7 Plus Preview5 of 7 tests · 1 tie
- AgentsCarrying out multi-step tasks on its own14Qwen3.7 Plus Preview4 of 5 tests
- ReasoningHard problems that need careful thinking03Qwen3.7 Plus Preview3 of 3 tests
- Images and chartsUnderstanding pictures, charts and video02Qwen3.7 Plus Preview2 of 4 tests · 2 ties
- MathCompetition and research-level math02Qwen3.7 Plus Preview2 of 2 tests
- Following instructionsDoing exactly what it is asked01Qwen3.7 Plus Preview1 of 1 test
- Long documentsFinding answers in very long texts10Kimi K2.51 of 1 test
- FactsGetting facts right instead of making them up11Even1 each
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 44% 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+38points ahead
- Command-line tasks in a real terminalTerminal-Bench 2.1+15.3points ahead
- Hard command-line tasks in a real terminalTerminal-Bench Hard+12.2points ahead
Where Kimi K2.5 pulls ahead
- Answers hard knowledge questions correctlyAA-Omniscience · Accuracy+12.7points ahead
- Reasons across sets of long documentsAA-LCR+5points ahead
- Real work tasks from 44 professionsGDPVal+3.6points ahead
Every test, side by side
All 41 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingQwen3.7 Plus Preview
- Terminal-Bench 2.1Qwen3.7 Plus Preview by 15.345.761+15.3
- Terminal-Bench HardQwen3.7 Plus Preview by 12.234.847+12.2
- SWE-bench ProQwen3.7 Plus Preview by 6.950.757.6+6.9
- LiveCodeBench v6Qwen3.7 Plus Preview by 4.68589.6+4.6
- SciCodeKimi K2.5 by 2.94946.1+2.9
- SWE-bench MultilingualQwen3.7 Plus Preview by 2.87375.8+2.8
- SWE-bench Verifiedtie76.877.7tie
AgentsQwen3.7 Plus Preview
- AA ApexAgentsQwen3.7 Plus Preview by 10.911.522.4+10.9
- OSWorld-VerifiedQwen3.7 Plus Preview by 1063.373.3+10
- MCP AtlasQwen3.7 Plus Preview by 9.463.873.2+9.4
- GDPValKimi K2.5 by 3.61713.4+3.6
- τ-Bench V3 · BankingQwen3.7 Plus Preview by 3.314.217.5+3.3
ReasoningQwen3.7 Plus Preview
- CritPtQwen3.7 Plus Preview by 63.19.1+6
- Humanity's Last ExamQwen3.7 Plus Preview by 4.930.735.6+4.9
- GPQA DiamondQwen3.7 Plus Preview by 2.187.990+2.1
FactsEven
- AA-Omniscience · Non-hallucinationQwen3.7 Plus Preview by 3834.372.3+38
- AA-Omniscience · AccuracyKimi K2.5 by 12.735.222.5+12.7
Images and chartsQwen3.7 Plus Preview
- CharXiv (RQ)Qwen3.7 Plus Preview by 8.477.585.9+8.4
- MMMU-ProQwen3.7 Plus Preview by 5.175.480.5+5.1
- LMArena · Visiontie12691277tie
- Video-MMEtie87.488tie
MathQwen3.7 Plus Preview
- HMMT Feb. 2026Qwen3.7 Plus Preview by 5.887.192.9+5.8
- IMOAnswerBenchQwen3.7 Plus Preview by 4.281.886+4.2
Following instructionsQwen3.7 Plus Preview
- IFBenchQwen3.7 Plus Preview by 7.870.278+7.8
Other results16 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.
- BabyVisionQwen3.7 Plus Preview by 33.936.570.4+33.9
- MCPMarkQwen3.7 Plus Preview by 29.229.558.7+29.2
- Terminal-Bench 2.0Qwen3.7 Plus Preview by 19.550.870.3+19.5
- WorldVQAQwen3.7 Plus Preview by 14.846.361.1+14.8
- SimpleVQAQwen3.7 Plus Preview by 10.571.281.7+10.5
- Artificial Analysis Coding IndexQwen3.7 Plus Preview by 9.146.855.9+9.1
- NL2RepoQwen3.7 Plus Preview by 9.13241.1+9.1
- AA-OmniscienceQwen3.7 Plus Preview by 8.4-7.31.1+8.4
- MATH-VisionQwen3.7 Plus Preview by 6.184.290.3+6.1
- τ²-Bench Telecom (AA run)Kimi K2.5 by 2.995.993+2.9
- OmniDocBench 1.5Qwen3.7 Plus Preview by 2.688.891.4+2.6
- AA Agentic IndexKimi K2.5 by 221.719.7+2
- AA IntelligenceQwen3.7 Plus Preview by 1.723.525.2+1.7
- MMLU-ProQwen3.7 Plus Preview by 1.487.188.5+1.4
- VideoMMMUKimi K2.5 by 1.286.685.4+1.2
- LVBenchtie75.976.2tie
Questions people ask
Which is better, Kimi K2.5 or Qwen3.7 Plus Preview?
Qwen3.7 Plus Preview wins six of the eight areas we test: coding, agents, reasoning, images and charts, math and following instructions. Kimi K2.5 wins long documents. They are level on facts.
Which is better for coding?
Qwen3.7 Plus Preview. It wins 5 of the 7 coding tests both models report; Kimi K2.5 wins 1, and 1 is a tie.
Which is cheaper?
Kimi K2.5 costs $0.60 per million input tokens and $3.00 per million output tokens; Qwen3.7 Plus Preview costs $0.40 and $1.60. That makes Qwen3.7 Plus Preview about 44% cheaper for the same work.
How do you compare the two?
We use the 41 benchmark tests both models have published scores on. The verdict counts the 25 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 16 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.