Grok 4.7 Fast drops in Cursor, Grok Build

// 1h agoMODEL RELEASE

Grok 4.7 Fast drops in Cursor, Grok Build

xAI has launched Grok 4.7 Fast, an infrastructure-accelerated model variant delivering roughly 2x output generation speed compared to the standard model. Currently available exclusively within Cursor and Grok Build at double standard token pricing, the model targets low-latency interactive coding while bypassing the public API and Grok Build's free tier.

// ANALYSIS

Doubling output speed directly tackles the biggest friction in agentic coding workflows, but charging a 100% token surcharge for infrastructure-only acceleration tests developer price sensitivity.

–Slashing round-trip latency in Cursor and Grok Build directly improves multi-turn agent execution, where waiting on token generation interrupts developer flow.
–A 2x price increase ($4/M input, $12/M output) without quality or context improvements is steep, especially as rival inference providers aggressively race to zero on token costs.
–Exclusivity to Cursor and Grok Build shows xAI prioritizing high-profile developer environments to showcase speed before scaling infrastructure to public API traffic.
–Disqualifying the fast tier from Grok Build's free credits indicates that high-throughput inference on frontier clusters remains economically constrained.

// TAGS

grok-4-7-fastcursorgrok-buildai-codingcoding-agentinferenceidellm

DISCOVERED

1h ago

2026-09-22

PUBLISHED

1h ago

2026-09-22

RELEVANCE

9/ 10

AUTHOR

XFreeze

// KEEP READING

More AI developer news from the feed

EXPLORE FULL FEED

NEWS2h ago

xAI ships three Grok models in nine weeks

xAI has dramatically shortened its frontier model shipping cycle, delivering Grok 4.5 on July 16, Grok 4.6 on August 12, and Grok 4.7 on September 21. Spanning barely nine weeks, each successive iteration has brought targeted capability upgrades, particularly pushing boundaries in code generation and reasoning.

UPDATE2h ago

Grok Bot upgrades to Grok 4.7

Grok Bot, xAI's agent platform that deploys persistent cloud-based AI teammates for autonomous workflows, is now powered by the newly released Grok 4.7 frontier model. The upgrade introduces a 500k-token context window and native harness training, strengthening multi-step task execution across browsers and terminals while maintaining existing serving speeds and pricing.

UPDATE3h ago

Grok Build streamlines subagent delegation, model consistency

Grok Build has received new updates designed to improve long-running agent workflows and background execution. Subagents can now automatically default to a general-purpose configuration without requiring explicit type declarations, inherit and lock in the parent model when a session begins, and maintain greater operational stability across complex multi-step tasks.