Deepgram vs OpenAI API

Deepgram uses usage-based pricing starting with $200 in free credit, pay-as-you-go rates per minute or per character, a Growth plan with annual pre-paid credits requiring a $4K/year minimum, and custom Enterprise pricing.

API inputs and outputs are not used for training by default, which differs from the consumer product. Retention periods and enterprise terms change, so read the current data usage policy rather than trusting a summary.

New users get $200 of free credit with no credit card required, which can be applied to speech-to-text, text-to-speech, or voice agent usage before any payment is needed.

No. The weights are not distributed. If self-hosting is a requirement, you are looking at open-weight models instead, with the operational and quality trade-offs that implies.

Usage is metered per minute of audio for speech-to-text and voice agent calls, and per 1,000 characters for text-to-speech, with add-ons like redaction and entity detection billed separately per minute.

Per token, with input and output priced differently and each model priced differently. Batch processing and cached input prefixes reduce it. The practical consequence is that your bill is a function of prompt design, not just of request count.

The same model family delivered by Microsoft under an Azure contract, with Azure identity, networking and regional controls, and a different release cadence for new models. Enterprises with an Azure agreement often choose it for procurement and data residency reasons rather than technical ones.

Cap input length, cache repeated prefixes, route easy requests to smaller models, use the batch path where latency does not matter, and set per-user limits before launch rather than after the first surprising invoice.