Compare AI models

Specifications
Parameters
2.8T>1T
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
Humanity's Last Exam · with tools
56%49.8%
GPQA Diamond
93.5%87.4%
Overview
CompanyMoonshot AIQwen
Release dateJul 16 2026Jan 26 2026
AccessOpen WeightClosed
Model detailsView modelView model

Frequently asked questions

Kimi K3 leads Qwen3-Max-Thinking on 2 of the 2 benchmarks they both report (Humanity's Last Exam, GPQA Diamond). 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-Max-Thinking. Qwen3-Max-Thinking shipped 171 days before Kimi K3, so benchmark comparisons should account for the intervening progress.

Kimi K3 has 2.8T parameters, while Qwen3-Max-Thinking has >1T. Context windows are 1M (Kimi K3) vs 256k (Qwen3-Max-Thinking). Kimi K3 is open weight, while Qwen3-Max-Thinking is closed.

On Humanity's Last Exam · with tools, Kimi K3 leads at 56% vs Qwen3-Max-Thinking at 49.8%. On GPQA Diamond, Kimi K3 leads at 93.5% vs Qwen3-Max-Thinking at 87.4%.