Kimi K2.5vsQwen2.5-Coder

Kimi K2.5
Qwen2.5-Coder
Specifications
Parameters
1T
32B
Context window
256k
128k
Benchmarks
Nonsense detection
BullshitBench v2
52%
Coding
SWE-Bench Verified
76.8%
Agentic coding
CursorBench v3.1
31.9%
Next.js coding
Next.js Evals
21%
Competitive coding
LiveCodeBench
85%
Science
GPQA Diamond
87.6%
Community preference (code)
Arena Elo (Code)
1433
Overview
CompanyMoonshot AIQwen
Release dateJan 27 2026Nov 12 2024
AccessOpen WeightOpen Weight

Which is better: Kimi K2.5 or Qwen2.5-Coder?

Kimi K2.5 and Qwen2.5-Coder don't publish scores on any of the same benchmarks, so there's no direct head-to-head comparison. Qwen2.5-Coder shipped 441 days before Kimi K2.5, so benchmark comparisons should account for the intervening progress.

Kimi K2.5 has 1T parameters, while Qwen2.5-Coder has 32B. Context windows are 256k (Kimi K2.5) vs 128k (Qwen2.5-Coder).

Direct benchmark comparisons are unavailable — Kimi K2.5 and Qwen2.5-Coder don't publish scores on any of the same benchmarks.

Frequently asked questions

Kimi K2.5 was released by Moonshot AI on Jan 27 2026.

Qwen2.5-Coder was released by Qwen on Nov 12 2024.

Kimi K2.5 has a 256k context window; Qwen2.5-Coder has 128k.

Other comparisons