Comparing 2 models

DeepSeek V3 vs Kimi K3

Facts first, then task-specific rankings. A highlighted cell is better only for that row, not universally.

metric
Model 1DeepSeek V3
Model 2Kimi K3
Best for
Best Local LLM for Coding (#1)
Code Completion (#1)
Provider
deepseek
moonshotai
Context
164K
1.0MBest
Input / 1M
$0.20Best
$3
Output / 1M
$0.80Best
$15
Tool use
yesBest
yesBest
Vision
yesBest
Reasoning
yesBest
Open-weight
yesBest
Released
Dec 1, 2024
Jul 16, 2026