GitHub[220K]Try OpenCode

Compare AI models: Kimi K2 Instruct vs Kimi K3

  1. Kimi K2 InstructMoonshotHollow points use fallbacks · hover for details
  2. Kimi K3MoonshotHollow points use fallbacks · hover for details
ReasoningCodingCost efficiencyContext windowMultimodalTool use
Normalized model capability scores
ModelReasoningCodingCost efficiencyContext windowMultimodalTool use
Kimi K2 Instruct0/100 — Reasoning not supported50/100 — No data; neutral placeholder49/10036/1000/10050/100 — Tool calling supported; no comparable benchmark
Kimi K3100/100100/1009/10092/10050/10050/100 — Tool calling supported; no comparable benchmark
Overview
Author
Context length
131K
1M
Reasoning
False
True
Input modalities
Text
Text, Image, Video
Output modalities
Text
Text
Providers
Pricing
Input
$0.40/ 1M
$3.30/ 1M
Output
$1.80/ 1M
$17/ 1M
Cached input
$0.20/ 1M
$0.33/ 1M
MomentumLast 2 mo
Unique users
No usage
243K
Completed sessions
No usage
776,358
Token share
No usage
0.05%
Tokens
No usage
554B-69%
No trend data
SEP 1OCT 26
RetentionWeek 1
Returning users
Pending
40.4%18K user-weeks

Related comparisons. Other model pairs to check.