ecobee · Rate Limits

Ecobee Rate Limits

ecobee does not publish a single numeric per-minute request quota. Instead its documentation gives explicit polling guidance and per-request bounds. Callers should poll for thermostat changes no more frequently than once every three minutes - using GET /thermostatSummary to detect revisions and only fetching full thermostat objects when a revision changes - and should keep no more than two to three concurrent HTTP requests open at a time. Reports carry hard per-request limits: the runtime report is capped at 25 thermostats and a 31-day period per request.

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

It captures 4 rate-limit definitions, measuring requests, connections, thermostats, and minutes.

The profile also includes 3 backoff/retry policies defined and response codes documented for throttled.

Tagged areas include Smart Home, Thermostat, IoT, Rate Limiting, and Polling.

4 Limits Throttle: 429
Smart HomeThermostatIoTRate LimitingPolling

Limits

Thermostat Polling Interval application
requests
no more than once per 3 minutes
Poll GET /thermostatSummary at most every three minutes; only fetch full objects when a revision changes.
Concurrent Requests application
connections
2-3 concurrent
Keep no more than two to three concurrent HTTP requests open at a time.
Runtime Report Size request
thermostats
25 thermostats, 31 days
A single runtime report request is limited to 25 thermostats and a maximum 31-day period.
Access Token Lifetime token
minutes
~60 minutes
Access tokens are short-lived and must be refreshed via the refresh_token grant.

Policies

Revision-Based Polling
Use the thermostatSummary revision list to detect changes efficiently instead of repeatedly requesting full thermostat objects.
Token Refresh
Refresh expired access tokens with grant_type=refresh_token rather than re-running the full authorization flow.
Backoff
On throttling or errors, implement exponential backoff with jitter and reduce polling frequency and concurrency.

Sources