Grok Automations: Persistent Agents—Schedule Free, Inbox Triggers on SuperGrok

Automations push consumer Grok from reactive chat toward persistent agents—the same category OpenAI and Anthropic are racing with work agents and scheduled tasks. Shipping schedule + inbox triggers on the free/paid split also clarifies xAI’s SuperGrok monetization: recurring agent labor, not only higher model caps.

xAI shipped Automations in Grok: users describe work once; Grok executes on a schedule or when an email trigger matches, then reports via email and/or app notification. Surfaces: grok.com/automations and the Grok apps on iOS and Android. Scheduled automations are available to everyone; email triggers require SuperGrok.

How it works

From the primary “Automations in Grok” post (x.ai/news/grok-automations):

• Instructions read like a chat message: tasks, attachments, connectors, skills, mode.
• Schedules: once, daily, weekdays, weekly, monthly, or yearly at a chosen local time.
• Email triggers: filter by sender, recipient, or subject; the matching message becomes run context.
• Each run opens a full conversation with history—users can continue the thread after Grok finishes.
• Creation paths: Automations UI, chat (“check the news every morning…”), or templates; pause/resume/edit/delete anytime; Run now for testing.

This is product surface on top of Grok models—not a new base-model release (Grok 4.5 remains the July 8 model story).

Limits

• Email triggers are SuperGrok-gated; schedule reliability and failure handling are product claims.
• Not a new model SKU—Automations quality tracks underlying Grok capability.
• Connector/skill depth for automated runs is not independently scored here.

Sources

• xAI: “Automations in Grok” (July 16, 2026)
• Product entry: grok.com/automations