#6
Grok 4.1 Thinking
grok-4.1-thinking

Grok 4.1 Thinking is a higher-effort variant of Grok 4.1 tuned for top-end benchmark performance and difficult reasoning tasks. It trades some latency for more deliberate multi-step thinking, which helps it reach leading results on creative writing, emotional intelligence, and general text quality evaluations. This mode is well-suited for in-depth analysis, complex planning, and other scenarios where answer quality matters more than raw speed, while still maintaining strong conversational helpfulness and empathy.

Performance Metrics

Helpfulness
Instruction Following
Comprehension
Empathy
Creative Writing
89.51
Helpfulness
91.5
Empathy
89.25
Instruction Following
94
Creative Writing
79.79
Comprehension
93
Speed
Avg 62 tok/s

Model Specifications

Release Date
November 17, 2025
Lab
xAI
Type
Proprietary
Context Size
256K
Max Output Tokens
8.2K
Cost per 1M tokens
$3.00 / $15.00
Model Inputs
Text, Images, Code, Web
Model Outputs
Text
Tool Calling
Enabled

Compare With Similar Models