Kimi K2.5 vs Qwen3.8 27B
Wins 0 of 7 areas
—
Wins 5 of 7 areas
Coding · Agents · Reasoning · Images and charts · Following instructions
Qwen3.8 27B is the stronger all-rounder.
Scores updated · 35 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.
- CodingWriting and fixing software15Qwen3.8 27B5 of 7 tests · 1 tie
- ReasoningHard problems that need careful thinking04Qwen3.8 27B4 of 4 tests
- AgentsCarrying out multi-step tasks on its own14Qwen3.8 27B4 of 5 tests
- Images and chartsUnderstanding pictures, charts and video01Qwen3.8 27B1 of 3 tests · 2 ties
- Following instructionsDoing exactly what it is asked01Qwen3.8 27B1 of 1 test
- FactsGetting facts right instead of making them up11Even1 each
- Long documentsFinding answers in very long texts11Even1 each
Following instructions 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 3% 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+35.4points ahead
- Command-line tasks in a real terminalTerminal-Bench 2.1+34.1points ahead
- Customer-service tasks in a simulated bankτ-Bench V3 · Banking+33.8points ahead
Where Kimi K2.5 pulls ahead
- Answers hard knowledge questions correctlyAA-Omniscience · Accuracy+19.6points ahead
- Finds hard-to-locate facts by browsing the webBrowseComp+13.3points ahead
- Questions about very long textsLongBench v2+8.7points 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.
CodingQwen3.8 27B
- Terminal-Bench 2.1Qwen3.8 27B by 34.145.779.8+34.1
- LMArena · WebDevQwen3.8 27B by 154 rating points14361590+154 rating
- SWE-bench ProQwen3.8 27B by 1150.761.7+11
- LiveCodeBench v6Qwen3.8 27B by 5.38590.3+5.3
- SciCodeKimi K2.5 by 2.44946.6+2.4
- SWE-bench VerifiedQwen3.8 27B by 176.877.8+1
- SWE-bench Multilingualtie7373.8tie
AgentsQwen3.8 27B
- τ-Bench V3 · BankingQwen3.8 27B by 33.814.248+33.8
- GDPValQwen3.8 27B by 291746+29
- OSWorld-VerifiedQwen3.8 27B by 2163.384.3+21
- MCP AtlasQwen3.8 27B by 13.563.877.3+13.5
- BrowseCompKimi K2.5 by 13.374.961.6+13.3
ReasoningQwen3.8 27B
- ARC-AGI-2Qwen3.8 27B by 30.611.842.4+30.6
- Humanity's Last ExamQwen3.8 27B by 3.230.733.9+3.2
- GPQA DiamondQwen3.8 27B by 2.687.990.5+2.6
- CritPtQwen3.8 27B by 2.33.15.4+2.3
FactsEven
- AA-Omniscience · Non-hallucinationQwen3.8 27B by 35.434.369.7+35.4
- AA-Omniscience · AccuracyKimi K2.5 by 19.635.215.6+19.6
Images and chartsQwen3.8 27B
- CharXiv (RQ)Qwen3.8 27B by 12.777.590.2+12.7
- MMMU-Protie75.476.3tie
- LMArena · Visiontie12691271tie
Long documentsEven
- LongBench v2Kimi K2.5 by 8.76152.3+8.7
- AA-LCRQwen3.8 27B by 47882+4
Following instructionsQwen3.8 27B
- IFBenchQwen3.8 27B by 9.370.279.5+9.3
Other results11 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.
- BabyVisionQwen3.8 27B by 49.136.585.6+49.1
- AA Agentic IndexQwen3.8 27B by 24.821.746.5+24.8
- ARC-AGI-1Qwen3.8 27B by 22.265.387.5+22.2
- Artificial Analysis Coding IndexQwen3.8 27B by 21.346.868.1+21.3
- MATH-VisionQwen3.8 27B by 10.484.294.6+10.4
- NL2RepoQwen3.8 27B by 10.33242.3+10.3
- AA IntelligenceQwen3.8 27B by 10.223.533.7+10.2
- LVBenchKimi K2.5 by 3.575.972.4+3.5
- AA-OmniscienceKimi K2.5 by 2.7-7.3-10+2.7
- DeepSearchQA (F1)Qwen3.8 27B by 2.677.179.7+2.6
- OmniDocBench 1.5Qwen3.8 27B by 2.388.891.1+2.3
Questions people ask
Which is better, Kimi K2.5 or Qwen3.8 27B?
Qwen3.8 27B wins five of the seven areas where both have results: coding, agents, reasoning, images and charts and following instructions. Kimi K2.5 wins none. They are level on facts and long documents.
Which is better for coding?
Qwen3.8 27B. 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; Qwen3.8 27B costs $0.50 and $3.00. That makes Qwen3.8 27B about 3% 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 24 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 11 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.