Warp: Native Grok Subscription Support for xAI Models in Warp Agent
WarpWarp added first-class, OAuth-based support for connecting an xAI Grok subscription directly to the Warp Agent, making Grok models natively available to users with SuperGrok ($30/month) or X Premium+ ($40/month). Activated via the /grok command, the integration stores OAuth tokens securely and refreshes them proactively. Previously, Grok was only reachable through a manual custom inference endpoint workaround.
Sources & Mentions
3 external resources covering this update
Native Grok comes to the Warp Agent
Warp has added first-class support for connecting an xAI Grok subscription directly to the Warp Agent. Users with a SuperGrok ($30/month) or X Premium+ ($40/month) subscription can now make Grok models natively available inside Warp via the /grok command.
How the integration works
The integration uses OAuth (bring-your-own-subscription) rather than the manual custom inference endpoint workaround that was previously the only option. After authenticating, Warp stores the OAuth tokens securely and refreshes them proactively in the background, so the connection stays live without manual re-authentication.
It is worth noting the distinction between layers: this brings Grok models into Warp's agent inference layer, not the separate Grok Build CLI. The supported tiers at launch are SuperGrok and X Premium+.
A long-requested feature
The community had been asking for native Grok support since 2025. The GitHub tracking issue (#5867) accumulated dozens of comments over roughly a year, and earlier the only path was a brittle custom endpoint configuration. This release delivers the first-class integration users were waiting for, putting Grok on equal footing with the other model providers Warp supports.