Rate Limits

Each API key can make a set number of requests a minute, depending on your plan:

Plan	Requests a minute per key
Free	30
Pro	120
Team	300
Business	600

The limit belongs to the key, so separate keys for separate jobs never slow each other down. Your keys follow the best plan among the organizations you belong to, and a new limit applies from the request after your plan changes.

Hosted MCP uses the same plan limits per browser-authorized connection or API key. Every authenticated MCP message consumes budget, including setup and refused requests.

Going Over the Limit

A request over the limit is refused with 429 Too Many Requests, the rate_limited error code and a Retry-After header giving the seconds to wait:

HTTP/1.1 429 Too Many Requests
Retry-After: 60
Content-Type: application/json

{ "error": "rate_limited", "message": "This key is over its plan's requests-a-minute limit." }

Wait at least Retry-After seconds before sending more. For the REST API, rate-limit refusals do not count as usage. MCP refusals appear in its request activity and error metrics.

Staying Under It

• Spread bursts out instead of sending them all at once.
• Cache answers that rarely change, such as GET /me.
• Watch Busiest Minute on the Metrics page to see how close each key comes to the limit.