grok-build-0.1 vs kimi-k2.7-code
Pricing, Performance & Features Comparison
xAI’s agentic coding model and the successor to grok-code-fast-1; also suited to inexpensive general tool-calling.
Input$1
Output$2
Latency (p50)5.7s
Output Limit–
Function Calling
JSON Mode
-
InputText, Image
OutputText
in$1out$2cache$0.2–
Latency (24h)
Success Rate (24h)
Moonshot’s coding- and agent-specialized model built on Kimi K2.6; thinking is always on.
Input$0.95
Output$4
Latency (p50)5.4s
Output Limit–
Function Calling
JSON Mode
-
InputText
OutputText
in$0.95out$4cache$0.19–