Grok 4 Fast Reasoning

JSON →
xai llm
text

A fast inference variant of Grok 4 optimized for reasoning tasks.

context window 2.0M tokens
max output 2.0M tokens
input price $0.2 / 1M tokens
output price $0.5 / 1M tokens
streamingreasoningcode-generationfunction-callingtool-useprompt-caching