Usage Limits

Raycast AI puts powerful models from Anthropic, OpenAI, Google, and more right at your fingertips. All AI usage through Raycast AI is subject to fair usage limits designed to keep the service fast, reliable, and available for everyone.

Free Request Limits

Mac V2 and Windows

All Raycast users on Mac V2 and on Windows receive free access to GPT-5.4 mini, subject to our standard usage limits. This means you can start using Raycast AI at no cost.

• 50 requests per minute
• 300 requests per hour

iOS and Mac V1

Free users on Raycast for Mac V1 and on iOS receive 50 free messages to try Raycast AI with any of the models available in Raycast Pro. Each message you send counts as one request. After these free messages are used, you can upgrade to a Raycast Pro plan, which includes a 14-day free trial.

Pro Request Limits

These limits are shared across all models within their group. As AI is a fast-moving space, we review our limits regularly and reserve the right to update these request limits at any time.

Raycast Pro Models

Raycast Pro gives you unlimited messages to dozens of AI models from several AI providers, including frontier models.

• 50 requests per minute
• 300 requests per hour

Exceptions: o3-mini, o4-mini — 150 requests per 24 hours.

Advanced AI Models

In addition to AI models and limits available in Raycast Pro, Advanced AI is a subscription add-on that provides higher request limits and leading frontier models.

• 75 requests per 3 hours
• 150 requests per 24 hours

Exceptions: o1, Claude 4 Opus, Claude 4.1 Opus, Claude 4.5 Opus, Claude 4.6 Opus — 50 requests per week per model.

Additional Information

• Context windows vary by model. More details about models available with Raycast AI can be found on the Raycast website.
• The Submit Without Generating action in Raycast V1's AI Chat does not count towards your request limits.
• If you are using Raycast AI and provided your own API key using Bring Your Own Key, the request limits mentioned above do not apply.