Kimi K2.5vsQwen3-Coder

Kimi K2.5
Qwen3-Coder
Specifications
Parameters
1T480B
Context window
256k256k
API pricing
Input price
$0.60$1.00
Output price
$3.00$5.00
Cached input price
$0.10
Cheapest input
$0.45SiliconFlow$0.22Google
Cheapest output
$2.25SiliconFlow$1.55Novita
Benchmarks
BullshitBench v2
52%20%
Benchmarks
SWE-Bench Verified
76.8%
Next.js Evals
16%
LiveCodeBench
85%
BrowseComp
60.6%
Humanity's Last Exam · with tools
30.1%
GPQA Diamond
87.6%
Overview
CompanyMoonshot AIQwen
Release dateJan 27 2026Jul 22 2025
AccessOpen WeightOpen Weight

Other comparisons

Kimi K2.5vsClaude Fable 5.1Qwen3-CodervsClaude Fable 5.1Kimi K2.5vsGPT-6 AstraQwen3-CodervsGPT-6 AstraKimi K2.5vsGemini 3.8 FlashQwen3-CodervsGemini 3.8 FlashKimi K2.5vsMuse Spark 1.3Qwen3-CodervsMuse Spark 1.3Kimi K2.5vsGrok 4.6Qwen3-CodervsGrok 4.6Kimi K2.5vsDeepSeek-V4.1-FlashQwen3-CodervsDeepSeek-V4.1-Flash

Frequently asked questions

Kimi K2.5 leads Qwen3-Coder on 1 of the 1 benchmark they both report (BullshitBench v2). Kimi K2.5 is cheaper on both input and output: $0.60 vs $1.00 per million input tokens, and $3.00 vs $5.00 per million output tokens. Figures are base-tier rates. Qwen3-Coder shipped 189 days before Kimi K2.5, so benchmark comparisons should account for the intervening progress.

Kimi K2.5 has 1T parameters, while Qwen3-Coder has 480B. Context windows are 256k (Kimi K2.5) vs 256k (Qwen3-Coder).

On BullshitBench v2, Kimi K2.5 leads at 52% vs Qwen3-Coder at 20%.