Raycast · Rate Limits

Raycast Rate Limits

Raycast's developer surface is the @raycast/api SDK rather than a metered HTTP API, so there is no public per-request RPM/TPM rate-limit table. The practical limits that exist are subscription quotas - the number of AI messages included per plan - and AI model fair-use limits enforced inside the app. SDK modules (Storage, Cache, OAuth, etc.) run locally and are not network rate-limited by Raycast. Specific AI fair-use values are not reconciled in this artifact.

Raycast Rate Limits is the machine-readable rate-limit profile for Raycast on the APIs.io network, conforming to the API Commons Rate Limits specification.

It captures 4 rate-limit definitions, measuring messages, requests, and calls.

The profile also includes 2 backoff/retry policies defined and response codes documented for throttled.

Tagged areas include Productivity, Launcher, Extensions, SDK, and AI.

4 Limits Throttle: n/a (SDK, not HTTP REST)
ProductivityLauncherExtensionsSDKAIRate LimitingQuotas

Limits

AI Messages (Free) account
messages
50 per month
Included AI message allowance on the Free plan; AI is accessed via the SDK AI.ask, not a public endpoint.
AI Messages (Pro / Teams) account
messages
see provider documentation
Paid plans raise AI allowances; Advanced AI add-on unlocks frontier models with higher fair-use limits.
AI Model Fair Use account
requests
fair-use, enforced in-app
Raycast applies fair-use throttling on AI model access; not exposed as documented per-request quotas.
SDK Module Calls device
calls
not network rate-limited
Storage, Cache, Clipboard, Preferences, OAuth and UI modules execute locally inside the extension process.

Policies

Subscription Quotas
AI usage is gated by plan tier (Free / Pro / Teams / Advanced AI) rather than per-request API limits.
Fair Use
AI model access is subject to in-app fair-use enforcement; abuse can be throttled at the account level.

Sources