Compare AI models

Specifications
Parameters
2.8T0.8B/2B/4B/9B
Context window
1M256k
API pricing
Input price
$3.00—
Output price
$15.00—
Cached input price
$0.30—
Cheapest input
$0.153Wafer—
Cheapest output
$9.50Chutes—
$535Wafer · mxfp4—
Benchmarks
GPQA Diamond
93.5%81.7%
OSWorld-Verified
84.8%41.8%
CharXiv Reasoning
84.8%73%
MMMU-Pro
81.6%70.1%
Overview
CompanyMoonshot AIQwen
Release dateJul 16 2026Mar 2 2026
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

Kimi K3 leads Qwen3.5 Small on 4 of the 4 benchmarks they both report (GPQA Diamond, OSWorld-Verified, CharXiv Reasoning, MMMU-Pro). Only Kimi K3 has a verified first-party price for an available API: $3.00 per million input tokens and $15.00 per million output tokens. No currently available pay-as-you-go API rate is tracked for Qwen3.5 Small. Qwen3.5 Small shipped 136 days before Kimi K3, so benchmark comparisons should account for the intervening progress.

Kimi K3 has 2.8T parameters, while Qwen3.5 Small has 0.8B/2B/4B/9B. Context windows are 1M (Kimi K3) vs 256k (Qwen3.5 Small).

On GPQA Diamond, Kimi K3 leads at 93.5% vs Qwen3.5 Small at 81.7%. On OSWorld-Verified, Kimi K3 leads at 84.8% vs Qwen3.5 Small at 41.8%. On CharXiv Reasoning, Kimi K3 leads at 84.8% vs Qwen3.5 Small at 73%. On MMMU-Pro, Kimi K3 leads at 81.6% vs Qwen3.5 Small at 70.1%.