
Deepgram
Speech-to-text, text-to-speech and voice agent APIs
Deepgram is a voice AI platform for developers whose speech-to-text API transcribes streaming and pre-recorded audio, with Nova models supporting 60+ languages. There is no free plan, but new accounts get a free $200 credit that does not expire. Pay-as-you-go transcription is listed from $0.0077 per minute on its pricing page. It suits teams adding transcription to their own apps.

What Deepgram does
Deepgram offers its speech recognition as an API. It transcribes both streaming and pre-recorded audio, and its Nova models support 60+ languages, with multilingual models that handle several languages in real time. Speaker diarization, Smart Formatting and Keyterm Prompting are available, and redaction and entity detection can be added on top of transcription for an extra per-minute fee.
The same platform also has a text-to-speech API, a Voice Agent API and audio intelligence features, so it covers more than transcription. A dedicated EU endpoint keeps processing in the European Union, and Enterprise customers can run self-hosted containers in their own VPC or on-prem. The site lists SOC 2 Type 1 and Type 2, HIPAA with a BAA for Enterprise, GDPR, CCPA and PCI compliance.
Billing is per second of audio, with no rounding up to a block and no minimum duration. Pay as you go starts with a free $200 credit, which the site says covers roughly 43,000 minutes of Nova-3 transcription, and then charges from $0.0077 per minute. Growth is pre-paid annual credits from $4,000 a year for up to 20% off those rates and more streaming concurrency. Enterprise is custom priced.
Features
- Speech-to-text API for streaming and pre-recorded audio
- Nova models supporting 60+ languages
- Speaker diarization
- Smart Formatting and Keyterm Prompting
- Redaction and entity detection add-ons
- Text-to-speech API (Flux TTS and Aura voices)
- Voice Agent API
- Audio intelligence features
- Dedicated EU endpoint for data residency
- Self-hosted deployment in your own VPC or on-prem (Enterprise)
Deepgram pricing
| Plan | Price | What's included (as stated) |
|---|---|---|
| Pay as you go | $0.0077 usagebilling term not confirmed |
|
| Growth | $4000/yrbilled annually |
|
| Enterprise | Custom |
|
Copied from Deepgram's pricing page on Nov 10, 2026. Prices change, so confirm before you buy.
Strengths and limits
Each point comes from the features and plan details on this page. No point is an opinion.
- StrengthA free $200 credit with no expiration and no credit card required.
- StrengthBilling runs per second of audio with no minimum duration; Nova-3 Monolingual is $0.0077 per minute on Pay as you go.
- StrengthHandles both streaming and pre-recorded audio, with Nova models supporting 60+ languages.
- StrengthEnterprise customers can self-host in their own VPC or on-prem.
- LimitNo ongoing free plan; the $200 credit is a one-off allowance.
- LimitRedaction and entity detection are optional add-ons rather than part of Nova speech-to-text.
- LimitThe lower Growth rates require pre-paying at least $4,000 a year.
What people use Deepgram for
Deepgram FAQ
Is Deepgram free?
There is no free plan, but new accounts get a free $200 credit that does not expire, with no credit card required.
How much does Deepgram cost?
Pay as you go transcription is listed from $0.0077 per minute. Growth is pre-paid annual credits from $4,000 a year for up to 20% off. Enterprise is custom priced.
Where this listing comes from
Facts were read from these pages on Nov 10, 2026. The description was written from those facts. See the methodology.
Something out of date? Report a correction. Corrections are free.
Deepgram alternatives

AssemblyAI
FreemiumUsage-based, from $0.15AI models to transcribe and understand speech
Transcription & MeetingsAudio & Voice
Rev
FreemiumFree plan availableInvestigative and legal transcription platform that turns case evidence into a searchable, citable record.
Transcription & MeetingsResearch & Data
Sonix
Free trialFrom $25/moAI transcription software for audio and video in 54+ languages, with translation, AI summaries and speaker detection
Transcription & Meetings