Compare AI models

Specifications
Parameters
2.8T27B
Context window
1M256k
API pricing
Input price
$2.00—
Output price
$10.00—
Cached input price
$0.20—
Cheapest input
$0.2403Relace—
Cheapest output
$6.00Decart—
$468Relace · fp4—
Benchmarks
Humanity's Last Exam · no tools
51.1%24%
GPQA Diamond
94.2%87.8%
CharXiv Reasoning
89%78.4%
MMMU-Pro
84.1%75.8%
Overview
CompanyMoonshot AIQwen
Release dateOct 23 2026Apr 22 2026
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

Kimi K3.1 Pro leads Qwen3.6-27B on 4 of the 4 benchmarks they both report (Humanity's Last Exam, GPQA Diamond, CharXiv Reasoning, MMMU-Pro). Only Kimi K3.1 Pro has a verified first-party price for an available API: $2.00 per million input tokens and $10.00 per million output tokens. No currently available pay-as-you-go API rate is tracked for Qwen3.6-27B. Qwen3.6-27B shipped 184 days before Kimi K3.1 Pro, so benchmark comparisons should account for the intervening progress.

Kimi K3.1 Pro has 2.8T parameters, while Qwen3.6-27B has 27B. Context windows are 1M (Kimi K3.1 Pro) vs 256k (Qwen3.6-27B).

On Humanity's Last Exam · no tools, Kimi K3.1 Pro leads at 51.1% vs Qwen3.6-27B at 24%. On GPQA Diamond, Kimi K3.1 Pro leads at 94.2% vs Qwen3.6-27B at 87.8%. On CharXiv Reasoning, Kimi K3.1 Pro leads at 89% vs Qwen3.6-27B at 78.4%. On MMMU-Pro, Kimi K3.1 Pro leads at 84.1% vs Qwen3.6-27B at 75.8%.