by Deepgram, Inc.Speech-to-text APIs
A speech-to-text API with Nova-3 batch transcription from $0.0043 a minute, speaker labels included on pre-recorded audio, and $200 of credit with no card.
Speaker-labelled transcripts of pre-recorded audio at $0.0043 a minute with nothing extra to add, plus regional endpoints and a self-hosting path for later.
On pay-as-you-go, Nova-3 Monolingual transcribes pre-recorded audio for $0.0043 a minute with speaker diarization, smart formatting and word timestamps included, billed per second. The $200 sign-up credit needs no card and covers about 775 hours of that workload. Teams that must keep audio in a region can use EU, Australia or India endpoints with the same keys, and Enterprise adds self-hosted containers for a VPC or on-prem GPUs.
Against 3 other products we reviewed in Speech-to-text APIs: what you pay for with each, the five access facts, the plan limits we recorded and each review's bottom line. The note under the table says how each one was checked.
| Metric | Deepgram | AssemblyAI | Speechmatics | Google Cloud Speech-to-Text |
|---|---|---|---|---|
| What you're paying for | Worth the priceSpeaker-labelled transcripts of pre-recorded audio at $0.0043 a minute with nothing extra to add, plus regional endpoints and a self-hosting path for later. | Compliance includedA transcription API whose BAA, training opt-out and retention controls are self-serve on paid accounts, with diarized batch audio from $0.17 an hour. | Compliance includedTranscription in 55+ languages where your audio stays out of model training unless you opt in, with EU, US and Australian processing for every account. | Built-in reachSpeech recognition inside your Google Cloud account, with IAM, audit logs, a BAA and EU processing, and dynamic batch at $0.003 a minute. |
| Overall score | 6.6 | 8.4Best | 8.1 | 7.8 |
| Free option | Free trial | Free trial | Free trial | Free plan |
| Card to start | No card needed | No card needed | No card needed | Card required |
| Getting access | Self-serve | Self-serve | Self-serve | Self-serve |
| Starter priceCheapest paid plan | $0.0043/min | $0.15/hr | $0.24/hr | $0.003/min |
| Pricing model | Usage-based | Usage-based | Usage-based | Usage-based |
| Pricing page | Public prices | Public prices | Public prices | Public prices |
| Billing unit | Per second of audio; multichannel billed per channel | Per second of pre-recorded audio; streaming by session time, idle included; multichannel per channel | Per second, from credits at hourly rates; credit price falls with packs and subscriptions | Per second of audio, no minimum; each channel billed |
| Free credit | $200 at sign-up, no card | $50 at sign-up, no card, does not expire | $100 at sign-up, no card, valid 12 months | 60 minutes a month on the V1 API; none on V2 |
| Streaming rate | Nova-3 Monolingual $0.0077/min regular; $0.0048/min limited-time promotional | Universal-Streaming $0.15/h; Universal-3.6 Pro Realtime $0.45/h | Melia 1 $0.40/h; Standard $0.45/h; Enhanced $0.80/h (list) | $0.016/min ($0.96/h), the same as batch |
| Diarization | Included on pre-recorded audio; streaming +$0.0020/min | Batch +$0.02/h; streaming +$0.12/h (beta in the docs) | Included; real-time speaker labels on Enhanced, Standard and Linden, not Melia 1 | No separate price; Chirp 3 diarization in batch only |
| Languages | 45+ (pricing page) | 99 on Universal-2; 18 on Universal-3.5 Pro | 55+ (pricing page); 56 for Melia 1 | Over 125 (product page) |
| Training on your audio | By default; opt out per request (mip_opt_out=true) | By default; paid accounts opt out online, free accounts cannot | Only if you opt in (33% off) | Not without your permission; V1 data logging opt-in gives a lower rate |
| HIPAA BAA | Enterprise, through sales | Signed online at no cost on paid plans | Sub-BAA through the Trust Center; plan and process not stated | Covered; accepted in the console per the docs |
| EU processing | EU endpoint with existing keys, no plan limit stated; also Australia and India | EU endpoint for batch and streaming, same price as US | EU, US and Australia regions for all customers | Chirp 3 GA in the eu multi-region |
| Self-hosted | Enterprise containers; Amazon SageMaker route with an AWS account | Streaming only, $20,000 upfront commitment | On-prem containers, virtual appliance and on-device on Enterprise | Speech-to-Text On-Prem on GKE, through sales |
| SOC 2 / ISO 27001 | SOC 2 Type II (report on request); ISO 27001 not stated | SOC 2 Type 2; ISO 27001:2022 | SOC 2 Type II (Trust Center, under NDA); ISO/IEC 27001:2022 | SOC 2 Type II and ISO 27001 (Compliance Reports Manager) |
| Console SSO | Not documented | SAML/OIDC add-on, $199/month per connection | SAML/OIDC on Enterprise, $280/month add-on | SAML through Cloud Identity Free, no listed cost |
| Bottom line | From its public pages, Deepgram pairs low batch rates, diarization included, with a $200 credit and self-hosting. Model improvement is on unless each request opts out, the BAA is Enterprise only, and streaming runs on promotional rates. | From its public pages, AssemblyAI combines the lowest 500-hour price here with compliance controls you switch on yourself: a BAA, a training opt-out and retention as short as an hour. A few add-on prices differ between its pages. | From its public pages, Speechmatics has the strictest privacy defaults of the five: no training unless you opt in, 7-day batch retention and real-time audio not stored. Its pricing page leads with discounted rates, and redaction, SSO and on-prem are Enterprise features. | From its public pages, Google Cloud Speech-to-Text suits teams inside Google Cloud: no training without permission, a BAA, EU processing and dynamic batch at $0.003 a minute. Features need checking per model, and there is no PII redaction. |
Evidence: Deepgram: facts checked 2026-10-16 from its public sign-up and pricing pages, no completed hands-on run; AssemblyAI: facts checked 2026-10-16 from its public sign-up and pricing pages, no completed hands-on run; Speechmatics: facts checked 2026-10-16 from its public sign-up and pricing pages, no completed hands-on run; Google Cloud Speech-to-Text: facts checked 2026-10-16 from its public sign-up and pricing pages, no completed hands-on run. Prices in USD as listed, excluding tax. "–" means not recorded.
Review from public pages
Written from Deepgram's public pages (pricing, security or privacy, and docs) as read on Oct 16, 2026. We did not sign up, and no hands-on run is published yet.
We read Deepgram's pricing page and its FAQ, the Terms of Service, the privacy policy, the data security page, and the developer docs on models and languages, diarization, redaction, keyterm prompting, multichannel, rate limits, self-hosting, data residency and the Model Improvement Program, plus the changelog, on Oct 16, 2026; we did not sign up or call the API for this review. The pricing page loaded in US dollars from our location.
The rate tables are detailed: per-minute prices for pay-as-you-go and Growth, model by model, for pre-recorded and streaming audio, with add-ons such as keyterm prompting and redaction priced the same way. Streaming rates are marked "Limited-time promotional rates on streaming" with no end date, so our tables use the regular rates; the FAQ quotes the promotional $0.29 an hour as the pay-as-you-go streaming price.
A few points need a question to Deepgram. The Growth discount is "up to 20%" in the plan header and "about 12.5%" in the FAQ. The pricing page says credits have no expiration, while the Terms end them when the account closes or a subscription is cancelled. Nova-3 Medical has no listed price, and the docs do not say whether streaming is billed on audio sent or on connection time.
$0.0043 a minute ($0.26 an hour) for Nova-3 Monolingual pre-recorded audio, with diarization included; our 500-hour scenario costs $129 a month (computed).
vs Amazon Transcribe: AssemblyAI adds $0.02 an hour for batch diarization and still comes to $85; Amazon Transcribe includes it at $0.006 a minute ($180 for the same scenario).
$200 of credit with no card, which the pricing page says has no expiration; at the scenario rate that is about 775 hours.
vs Amazon Transcribe: AssemblyAI gives $50 and Speechmatics $100; Google Cloud and Amazon Transcribe offer 60 free minutes a month.
EU, Australia and India endpoints work with existing API keys, with no plan limit stated; the docs say a full in-region guarantee needs the regional endpoint plus mip_opt_out=true.
Enterprise customers can run Deepgram in containers in their own VPC or on on-prem NVIDIA GPUs; the docs also describe an Amazon SageMaker deployment that needs only an AWS account.
vs Amazon Transcribe: AssemblyAI self-hosts streaming only, with a $20,000 commitment; Amazon Transcribe has no self-hosted option.
The Terms let Deepgram use your content to train its models, and requests join the Model Improvement Program unless each one sets mip_opt_out=true; opted-out requests get zero retention.
Workaround: Since March 5, 2026, opting out has no effect on listed prices on pay-as-you-go and Growth.
The HIPAA BAA is for Enterprise customers, and the privacy policy says DPAs are entered into with customers; we found no DPA that applies with the terms. ISO 27001 and console SSO are not stated.
Workaround: The SOC 2 Type II report is available from [email protected] or your account executive.
Streaming rates are labelled limited-time promotional with no end date; the regular Nova-3 Monolingual streaming rate is $0.0077 a minute against $0.0048 promotional.
Workaround: Budget on the regular rates shown in the same table.
Streaming diarization and redaction are $0.0020 a minute each and keyterm prompting $0.0013; entity redaction is English only, and multichannel audio is billed per channel.
Deepgram bills each second of audio from prepaid credit. For our scenario, 500 hours of pre-recorded English with diarization and word timestamps on Nova-3 Monolingual cost 30,000 × $0.0043 = $129 a month (computed) before tax, since diarization is included on pre-recorded audio. Keyterm prompting adds $0.0013 a minute and redaction $0.0020 ($39 and $60 a month at that volume, computed). Streaming costs more than batch: $0.0077 a minute regular for Nova-3 Monolingual, plus $0.0020 for diarization. Growth prepays $4,000 or more a year for up to 20% off, with overage at the Growth rate plus 10%; larger volume discounts are on Enterprise.
Transcripts come back as JSON, with SRT and WebVTT through the SDK caption helpers. The request-history API lists past requests but not their audio or transcripts, so keep your own copies. The FAQ offers refunds for credits bought in the last 30 days; the Terms otherwise say unused credits are not refunded. Growth renews yearly unless cancelled at least a day before the term ends.
From its public pages, Deepgram pairs low batch rates, diarization included, with a $200 credit and self-hosting. Model improvement is on unless each request opts out, the BAA is Enterprise only, and streaming runs on promotional rates.
| Limit | As recorded |
|---|---|
| Billing unit | Per second of audio; multichannel billed per channel |
| Free credit | $200 at sign-up, no card |
| Streaming rate | Nova-3 Monolingual $0.0077/min regular; $0.0048/min limited-time promotional |
| Diarization | Included on pre-recorded audio; streaming +$0.0020/min |
| Languages | 45+ (pricing page) |
| Training on your audio | By default; opt out per request (mip_opt_out=true) |
| Retention | Kept for model improvement by default; zero retention with the opt-out |
| HIPAA BAA | Enterprise, through sales |
| EU processing | EU endpoint with existing keys, no plan limit stated; also Australia and India |
| Self-hosted | Enterprise containers; Amazon SageMaker route with an AWS account |
| SOC 2 / ISO 27001 | SOC 2 Type II (report on request); ISO 27001 not stated |
| Console SSO | Not documented |
Read from the vendor's public pages on Oct 16, 2026. No hands-on run is published yet. How we check
We have not published a sign-up run for Deepgram yet. The access facts above were read from its sign-up and pricing pages on Oct 16, 2026.
We have not published API calls for Deepgram yet.
Deepgram lists 6 plans. The cheapest paid plan, Nova-3 Monolingual, pre-recorded (pay-as-you-go), costs $0.0043/min.
Prices as listed, USD, excluding tax; observed 2026-10-16.
Source: deepgram.com pricing page. Prices change; the observed date above is when our agent read them.
We read the pricing, docs, data-usage, trust and legal pages of three transcription APIs on the same day and scored them on the same six indicators. What 500 hours of diarized batch audio cost, how each handles your audio, where the BAA sits, and who should pick which.
