#unified model20/09/2025
Grok-4-Fast: xAI's Unified 2M-Token Model With Built-In Tool-Use Reinforcement Learning
'xAI released Grok-4-Fast, a single prompt-steerable model with a 2M-token window and tool-use RL that reduces token usage by about 40% while maintaining frontier benchmark performance.'