Kimi K2 ThinkingvsQwen3.5

Kimi K2 Thinking
Qwen3.5
Specifications
Parameters
1T
397B
Context window
256k
1M
Benchmarks
Coding
SWE-Bench Verified
71.3%
76.4%Best
Multilingual coding
SWE-Bench Multilingual
69.3%
Agentic terminal coding
Terminal-Bench 2.0
52.5%
Web browsing
BrowseComp
69%
Multidisciplinary reasoning
Humanity's Last Exam · no tools
28.7%
Science
GPQA Diamond
88.4%
Agentic computer use
OSWorld-Verified
62.2%
Chart reasoning
CharXiv Reasoning
80.8%
Multimodal reasoning
MMMU-Pro
79%
Multimodal
MMMU
85%
Overview
CompanyMoonshot AIQwen
Release dateNov 6 2025Feb 16 2026
AccessOpen WeightOpen Weight

Which is better: Kimi K2 Thinking or Qwen3.5?

Qwen3.5 leads Kimi K2 Thinking on 1 of the 1 benchmark they both report (SWE-Bench Verified). Kimi K2 Thinking shipped 102 days before Qwen3.5, so benchmark comparisons should account for the intervening progress.

Kimi K2 Thinking has 1T parameters, while Qwen3.5 has 397B. Context windows are 256k (Kimi K2 Thinking) vs 1M (Qwen3.5).

On SWE-Bench Verified, Qwen3.5 leads at 76.4% vs Kimi K2 Thinking at 71.3%.

Frequently asked questions

Kimi K2 Thinking was released by Moonshot AI on Nov 6 2025.

Qwen3.5 was released by Qwen on Feb 16 2026.

Qwen3.5 leads on SWE-Bench Verified — Kimi K2 Thinking 71.3% vs Qwen3.5 76.4%.

Kimi K2 Thinking has a 256k context window; Qwen3.5 has 1M.

Other comparisons