Compare AI models

Specifications
Parameters
1T27B
Context window
256k256k
API pricing
Cheapest input
$0.49AtlasCloud—
Cheapest output
$2.50AtlasCloud—
$223AtlasCloud · int4—
Benchmarks
SWE-Bench Verified
76.8%77.2%
SWE-Bench Multilingual
73%71.3%
LiveCodeBench
85%83.9%
Terminal-Bench 2.0
50.8%59.3%
Humanity's Last Exam · no tools
30.1%24%
GPQA Diamond
87.6%87.8%
MMLU-Pro
87.1%86.2%
Overview
CompanyMoonshot AIQwen
Release dateJan 27 2026Apr 22 2026
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

Kimi K2.5 leads Qwen3.6-27B on 4 of the 7 benchmarks they both report. Kimi K2.5 shipped 85 days before Qwen3.6-27B, so benchmark comparisons should account for the intervening progress.

Kimi K2.5 has 1T parameters, while Qwen3.6-27B has 27B. Context windows are 256k (Kimi K2.5) vs 256k (Qwen3.6-27B).

On SWE-Bench Verified, Qwen3.6-27B leads at 77.2% vs Kimi K2.5 at 76.8%. On SWE-Bench Multilingual, Kimi K2.5 leads at 73% vs Qwen3.6-27B at 71.3%. On LiveCodeBench, Kimi K2.5 leads at 85% vs Qwen3.6-27B at 83.9%. On Terminal-Bench 2.0, Qwen3.6-27B leads at 59.3% vs Kimi K2.5 at 50.8%. On Humanity's Last Exam · no tools, Kimi K2.5 leads at 30.1% vs Qwen3.6-27B at 24%. On GPQA Diamond, Qwen3.6-27B leads at 87.8% vs Kimi K2.5 at 87.6%. On MMLU-Pro, Kimi K2.5 leads at 87.1% vs Qwen3.6-27B at 86.2%.