Compare AI models

Specifications
Parameters
671B1T
Context window
128k256k
API pricing
Input price
—$0.95
Output price
—$4.00
Cached input price
—$0.16
Cheapest input
$0.256DeepInfra$0.3895Baidu
Cheapest output
$0.712DeepInfra$1.64Baidu
$103StreamLake$160Baidu · fp4
Benchmarks
SWE-Bench Verified
42%80.2%
LiveCodeBench
40.5%89.6%
GPQA Diamond
59.1%90.5%
Overview
CompanyDeepSeekMoonshot AI
Release dateDec 26 2024Apr 20 2026
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

Kimi K2.6 leads DeepSeek V3 on 3 of the 3 benchmarks they both report (SWE-Bench Verified, LiveCodeBench, GPQA Diamond). Only Kimi K2.6 has a verified first-party price for an available API: $0.95 per million input tokens and $4.00 per million output tokens. No currently available pay-as-you-go API rate is tracked for DeepSeek V3. DeepSeek V3 shipped 480 days before Kimi K2.6, so benchmark comparisons should account for the intervening progress.

DeepSeek V3 has 671B parameters, while Kimi K2.6 has 1T. Context windows are 128k (DeepSeek V3) vs 256k (Kimi K2.6).

On SWE-Bench Verified, Kimi K2.6 leads at 80.2% vs DeepSeek V3 at 42%. On LiveCodeBench, Kimi K2.6 leads at 89.6% vs DeepSeek V3 at 40.5%. On GPQA Diamond, Kimi K2.6 leads at 90.5% vs DeepSeek V3 at 59.1%.