Status

The measured success rate and generation time of every model. Every number is measured on this service after launch. There are no estimates or targets.

The measurements cannot be read right now. Try again in a few minutes.

All models

The success rate across every model.

PeriodSuccess rateResults measured
Last 24 hours——
Last 7 days——
Last 30 days——

By model

The success rate and generation time of each model. Measuring means that not enough has been measured yet: fewer than 20 results for the success rate, or fewer than 20 successful results for the generation time.

CherryText 1.0 Flash Text

PeriodSuccess rateGeneration time (median)Generation time (95th percentile)Results measured
Last 24 hours————
Last 7 days————
Last 30 days————

Cherry 1.0 Pro Video

PeriodSuccess rateGeneration time (median)Generation time (95th percentile)Results measured
Last 24 hours————
Last 7 days————
Last 30 days————

CherryImage 1.0 Pro Image

PeriodSuccess rateGeneration time (median)Generation time (95th percentile)Results measured
Last 24 hours————
Last 7 days————
Last 30 days————

How the numbers are measured

  • A result fails when it ends with a 5xx status. A 4xx result, such as an invalid request or a content policy rejection, counts as a success. An image or video request that is rejected before its job is created is not counted.
  • Generation time is measured from successful results only. For a text model it is the time from the request to the start of the response (for a stream, until the response starts; the time to send the content is not included). For an image or video model it is the time from the creation of the job to its completion.
  • A success rate is shown only after 20 or more results have been measured, and a generation time only after 20 or more successful results have been measured. Until then the cell says Measuring.
  • The measurements are read from our logs and refreshed about every 5 minutes.
  • Every number is measured on this service after launch. There are no estimates or targets.