Control Grok model reasoning effort with reasoning_effort parameter
Adjust how much effort Grok models spend thinking before responding using the reasoning_effort parameter.
reasoning_effort
Where
API request parameters
Why
To fine-tune the model's thinking process for complex tasks, balancing response time and depth of analysis.
What it does and when to use it
The reasoning_effort parameter in Grok models controls the computational resources dedicated to the model's internal thinking process before generating a response. It can be set to 'low', 'medium', 'high', or 'xhigh' (on supported models like grok-4.6 and later). The default is 'high'.
Sources and credit
- Reasoning ↗xAI · xAI Docs: Guides · Aug 11, 2026 · paraphrased
First seen on Subvolts Aug 11, 2026. Updated Sep 11, 2026. Spotted a problem? Tell us.
Where it shows up
- ReasoningxAI · xAI Docs: Guides · Aug 11, 2026
Related in the PlaybookPlaybook →
The Playbook: reusable how-tos, prompts, configs and tips pulled out of the news.
- Setting tipManually override request timeout for reasoning models with streamingAdjust request timeout when streaming with reasoning models to prevent premature connection closure. · 5 min
- HeuristicProvide Clear Instructions for Code ExecutionGive detailed instructions, specify data formats and constraints, and provide context for code execution. · 2 min
- Code snippetStream Grok model's reasoning summary deltas alongside responseExample code for streaming reasoning summary deltas with the final response from Grok-4.6. · 5 min
- Setting tipUse reasoning.encrypted_content for conversational contextPass reasoning.encrypted_content to the Responses API to provide context from previous conversations. · 2 min
- ToolxAI Collections Search tool for Grok knowledge base integrationEnables Grok to search uploaded knowledge bases for accurate, contextually relevant responses. · 10 min
- ToolxAI Grok Code Execution Tool for PythonEnables Grok to write and execute Python code for calculations, data analysis, and math. · 5 min
Questions people ask
- Which AI tools does this work with?
- The sources used it with Grok.
- Where did this come from?
- xAI (xAI Docs: Guides). Subvolts collected, normalized and credited it; the link above opens the original.