The seat-based enterprise plan comes with a specified number of transcription hours per editor per year, which are shared across all editors in your workspace. For example, if you have 5 editors on a plan that includes "240hrs per editor", your team will have a pool of 1,200 transcription hours for the year. You can use the hours anytime during the year, as your needs fluctuate. Additionally, we don't charge for long periods of silence detected in your audio, ensuring you're only billed for active content. In fact, for one of our largest public defender clients, who work with a lot of body cam footage, about 36% of the audio uploaded is silent and not charged.

If you exceed the transcription limit for your plan, you'll be charged an overage fee per minute of video uploaded. For utilization-based model, you will be charged per minute of transcription, based on your usage.

All data is encrypted in transit and rest, and we have extremely high standards for data processing. Our platform runs on Google Cloud, which provides robust security features, and we are SOC 2 Type 2 certified. This means we undergo annual audits to ensure we meet stringent security standards. We have regular data backups to safeguard against any potential hardware failures or other issues. Your data is securely stored and can be restored if necessary.

No language models from subprocessors are required to use Reduct. We do use OpenAI's GPT models for certain features, such as summarization; these features can be turned off if desired. We also have a signed agreement with OpenAI in which OpenAI promises not to train on any data that we submit to them using the API, and that all data they receive from us is deleted within 30 days.