Grok 4.6

Grok 4.6 is SpaceXAI's flagship model for long-running coding, agentic knowledge work, and ambitious interactive or visual projects.

Language ModelCoding AssistantAgent WorkflowsVision InputTool UseSpaceXAI

SpaceXAI API / Grok Build / Cursor / GitHub Copilot / model gateways

Below 200K prompt tokens: $2/1M input, $0.50/1M cached input, $6/1M output; above 200K: $4/$1/$12; fast variant 2x

Commercial

🚀Function Overview

Grok 4.6 is a generally available frontier-value candidate with moderate token prices, a 500K context window, broad partner access, and a focus on sustained agent work.

Key Features

Official SpaceXAI model ID grok-4.6
500,000-token context window and no documented text-output limit
Text and image input with text output
Low, medium, high default, and xhigh reasoning effort
January 2026 knowledge cutoff
Responses API and Chat Completions support
Function calling, web search, X search, and code execution
Base pricing below 200K prompt tokens: $2 input, $0.50 cached input, and $6 output per million tokens
Pricing above 200K prompt tokens: $4 input, $1 cached input, and $12 output per million tokens
Available through the SpaceXAI API, Grok Build, Cursor, GitHub Copilot, and supported model gateways

Use Cases

•Long-running coding and repository agents
•Interactive and visual application prototyping
•Knowledge work using web, X search, or code execution
•Frontier-value routing across API and coding-product surfaces
•High-context tasks with explicit long-prompt cost controls

⚙️Input Parameters

model

string

Use grok-4.6 through a supported SpaceXAI or partner model surface.

input

object

Text or supported image input sent through the Responses API or equivalent supported client.

reasoning_effort

string

Choose low, medium, high, or xhigh; high is the documented default.

tools

array

Optional function, web search, X search, or code-execution tools supported by the active surface.

💡Usage Examples

Example 1

Input Parameters

{
  "model": "grok-4.6",
  "input": "Investigate this codebase, implement the requested feature, and verify it against the existing tests.",
  "reasoning_effort": "high"
}

Output Results

Grok 4.6 returns a text response through the supported API surface. Measure cache hits, task success, token use, and long-prompt pricing before production routing.

Quick Actions

Technical Specifications

Hardware Type: SpaceXAI API / Grok Build / Cursor / GitHub Copilot / model gateways
Commercial Use: Supported
Pricing: Below 200K prompt tokens: $2/1M input, $0.50/1M cached input, $6/1M output; above 200K: $4/$1/$12; fast variant 2x