Grok 4.3vsKimi K2.7 Code
Kimi K2.7 Code is cheaper than Grok 4.3 at $0.95/M vs $1.25/M input tokens.
| Spec | Grok 4.3 | Kimi K2.7 Code |
|---|---|---|
| Input price | $1.25/M tokens | $0.95/M tokens |
| Output price | $2.5/M tokens | $4/M tokens |
| Context window | 1.0M tokens | 262K tokens |
| Weights | Closed | Open |
| Free API (OpenRouter) | No | No |
| Released | May 2026 | Jun 2026 |
Our Verdict
Grok 4.3
Kimi K2.7 Code
Not enough votes to call it. On the specs, nothing separates them.
Too close to call
Reviewing agent-written code?See a Brief PR report
API pricing
Cost per 1M tokens
Grok 4.3
Input
$1.25
Output
$2.50
1.6× cheaper
Kimi K2.7 Code
Input
$0.95
1.3× cheaper
Output
$4.00
Kimi K2.7 Code wins input (1.3× cheaper)·Grok 4.3 wins output (1.6× cheaper)
Where to run it
15 hosts, cheapest first
Grok 4.31 host
Kimi K2.7 Code14 hosts
DeepInfrafp4$0.68 in$3.40 out262k99.8% upInceptronint4$0.71 in$3.21 out262k99.6% upCoreWeaveint4$0.71 in$3.50 out262k100% upStreamLake$0.71 in$3.00 out256k99.6% upVeniceint4$0.75 in$3.50 out256k98.5% upModelRunfp4$0.85 in$3.75 out262k100% up
8 more hostsFewer hosts
SiliconFlowfp8$0.86 in$3.80 out262k99.9% upNovitaint4$0.91 in$3.84 out262k100% upAlibaba Cloudfp8$0.95 in$4.00 out262k99.9% upBasetenfp4$0.95 in$4.00 out262k99.9% upCloudflare Workers AI$0.95 in$4.00 out262k98.8% upFireworks$0.95 in$4.00 out262k94.2% upGMI Cloudfp8$0.95 in$4.00 out262k99.7% upMoonshot AIint4$0.95 in$4.00 out262k100% up
Per million tokens. Prices and uptime via OpenRouter, checked 11 Sep 2026.
Writing DNA
Style Comparison
Similarity
63%
Grok 4.3 uses 2.7x more transitions
Grok 4.3
Kimi K2.7 Code
62%Vocabulary57%
16wSentence Length19w
0.46Hedging0.51
2.3Bold4.7
2.5Lists3.3
0.00Emoji0.00
0.34Headings0.76
0.14Transitions0.05
Based on 27 + 27 text responses
Research
What we learned reading every model
FAQ
Common questions
Keep exploring
