grok-4.6

Common Name: Grok 4.6

Released on Aug 12 12:00 AMSupportedTool InvocationSupportedReasoning

xAI's latest flagship reasoning model, focused on long-running agents and multimodal chat, tool use, and structured outputs.

Context

500K

Maximum Output

128K

Inputtext, image

Outputtext

Collecting…

< 200K

Input$2.20/MTokens

Cached Input$0.55/MTokens

Output$6.60/MTokens

200K+

Input$4.40/MTokens

Cached Input$1.10/MTokens

Output$13.20/MTokens

Performance Metrics (24h)

$1.10/$2.20/M

ctx256Kmax128Kavail—tps—

InOutCap

xAI's coding model for agentic software engineering and workflow tasks.

$2.20/$6.60/M

ctx500Kmax128Kavail—tps—

InOutCap

xAI's latest flagship reasoning model for multimodal chat, tool use, and structured outputs.

$1.375/$2.75/M

ctx1.0Mmax128Kavail—tps—

InOutCap

xAI's flagship reasoning model for agentic tool calling, instruction following, and multimodal text generation.

$1.375/$2.75/M

ctx1.0Mmax128Kavail—tps—

InOutCap

A reasoning-enabled Grok 4.20 snapshot for long-context multimodal problem solving.