by Speechmatics (Cantab Research Ltd)Speech-to-text APIs
A speech-to-text API with Melia 1 batch transcription at $0.24 an hour list price, audio kept out of training unless you opt in, and $100 of free credit.
Transcription in 55+ languages where your audio stays out of model training unless you opt in, with EU, US and Australian processing for every account.
The Terms say Speechmatics does not use your inputs or outputs for model training unless you opt in from the portal; opting in takes 33% off. Batch files are deleted after 7 days, or sooner through the API, and real-time audio is not stored. Every account can choose the EU, US or Australia region. Melia 1 transcribes 56 languages for $0.24 an hour at list price with diarization included, and Enterprise adds on-prem, virtual appliance and on-device deployment.
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 | Speechmatics | AssemblyAI | Google Cloud Speech-to-Text | Amazon Transcribe |
|---|---|---|---|---|
| What you're paying for | 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. | 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. | 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. | Built-in reachTranscription inside your AWS account at $0.006 a minute for batch, with speaker labels, custom vocabulary and language identification in the rate. |
| Overall score | 8.1 | 8.4Best | 7.8 | 7.0 |
| Free option | Free trial | Free trial | Free plan | Free trial |
| Card to start | No card needed | No card needed | Card required | –Not measured |
| Getting access | Self-serve | Self-serve | Self-serve | Self-serve |
| Starter priceCheapest paid plan | $0.24/hr | $0.15/hr | $0.003/min | $0.006/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, from credits at hourly rates; credit price falls with packs and subscriptions | Per second of pre-recorded audio; streaming by session time, idle included; multichannel per channel | Per second of audio, no minimum; each channel billed | Per second, no minimum; two channels at no extra charge |
| Free credit | $100 at sign-up, no card, valid 12 months | $50 at sign-up, no card, does not expire | 60 minutes a month on the V1 API; none on V2 | 60 minutes a month for 12 months |
| Streaming rate | Melia 1 $0.40/h; Standard $0.45/h; Enhanced $0.80/h (list) | Universal-Streaming $0.15/h; Universal-3.6 Pro Realtime $0.45/h | $0.016/min ($0.96/h), the same as batch | $0.01/min ($0.60/h) |
| Diarization | Included; real-time speaker labels on Enhanced, Standard and Linden, not Melia 1 | Batch +$0.02/h; streaming +$0.12/h (beta in the docs) | No separate price; Chirp 3 diarization in batch only | Included in batch and streaming, up to 30 speakers |
| Languages | 55+ (pricing page); 56 for Melia 1 | 99 on Universal-2; 18 on Universal-3.5 Pro | Over 125 (product page) | About 90 languages across 113 language rows (our count) |
| Training on your audio | Only if you opt in (33% off) | By default; paid accounts opt out online, free accounts cannot | Not without your permission; V1 data logging opt-in gives a lower rate | By default; opt out with an AWS Organizations policy |
| HIPAA BAA | Sub-BAA through the Trust Center; plan and process not stated | Signed online at no cost on paid plans | Covered; accepted in the console per the docs | Accepted in AWS Artifact |
| EU processing | EU, US and Australia regions for all customers | EU endpoint for batch and streaming, same price as US | Chirp 3 GA in the eu multi-region | Six EU regions for batch, four for streaming |
| Self-hosted | On-prem containers, virtual appliance and on-device on Enterprise | Streaming only, $20,000 upfront commitment | Speech-to-Text On-Prem on GKE, through sales | Not offered |
| SOC 2 / ISO 27001 | SOC 2 Type II (Trust Center, under NDA); ISO/IEC 27001:2022 | SOC 2 Type 2; ISO 27001:2022 | SOC 2 Type II and ISO 27001 (Compliance Reports Manager) | SOC 1, 2 and 3 and ISO 27001 (AWS Artifact) |
| Console SSO | SAML/OIDC on Enterprise, $280/month add-on | SAML/OIDC add-on, $199/month per connection | SAML through Cloud Identity Free, no listed cost | IAM Identity Center, no extra charge |
| Bottom line | 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, 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, 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. | From its public pages, Amazon Transcribe is the simple choice inside AWS: flat per-minute rates with the main features included and self-serve compliance paperwork. Voice inputs are used to improve the service unless your organization opts out. |
Evidence: Speechmatics: 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; Google Cloud Speech-to-Text: facts checked 2026-10-16 from its public sign-up and pricing pages, no completed hands-on run; Amazon Transcribe: 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 Speechmatics'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 Speechmatics' pricing page and its markdown copy, the docs on plans, billing, regions, limits, models, languages, diarization, custom dictionary, security and compliance, and SSO, the Terms of Service, the model training addendum, the security page, the enterprise page and the brand page on Oct 16, 2026; we did not sign up or call the API for this review. The Trust Center returned an error from our location, so we read its Internet Archive copy of Dec 3, 2025.
The pricing page opens on discounted rates: the headline "from $0.121/hr" for Melia 1 already takes off 33% for opting in to model training and 25% for the top annual subscription. The list rate that pay-as-you-go customers pay is $0.24 an hour, and we used list rates throughout.
Some pages need a question to Speechmatics. The docs describe a volume discount above 500 hours a month for each product type without a rate, while the pricing page says pay-as-you-go has no discounts. Oak 1 has a real-time price but is marked "Coming soon", and the docs call it batch only. The Terms still describe monthly free hours from an older plan.
Training on customer data happens only if you opt in from the portal; the 33% discount is the incentive, and the default is off.
vs AssemblyAI: Deepgram, AssemblyAI and Amazon Transcribe use customer audio by default unless you opt out.
Speaker diarization is an included feature: in batch on every model, in real-time on Enhanced, Standard and Linden.
vs AssemblyAI: Deepgram and AssemblyAI charge $0.12 an hour for streaming diarization.
EU, US and Australia regions are open to all customers for batch and real-time, and batch files are deleted after 7 days.
Enterprise adds on-prem containers, a virtual appliance, hybrid and on-device deployment.
vs AssemblyAI: Amazon Transcribe is cloud only; AssemblyAI self-hosts streaming only.
The pricing page shows rates after a 33% training opt-in and a 25% annual commitment; the pay-as-you-go list rate for Melia 1 is about twice the headline ($0.24 against $0.121 an hour).
Workaround: The breakdown at the top of the pricing page shows the $0.24 base rate and each discount.
Melia 1, the multilingual model the pricing page leads with, does not yet have real-time speaker labels, custom dictionary or confidence scores, and translation and summaries run on Enhanced and Standard only.
Workaround: Enhanced and Standard include these, at $0.75 and $0.45 an hour for batch.
PII redaction is "On request" for Enterprise, SSO is a $280 a month add-on on Enterprise, and the BAA process is not described beyond a Sub-BAA in the Trust Center.
The $100 sign-up credit, credit packs and annual subscription credits last 12 months; monthly subscription credits last 2 months.
Speechmatics bills to the second from credits; each model uses credits at a fixed hourly rate, and the price of a credit depends on how you pay. For our scenario, 500 hours of pre-recorded English with diarization and word timestamps on Melia 1 cost 500 × $0.24 = $120 a month at list price (computed) before tax; diarization is included, and the docs' volume discount applies only above 500 hours. Opting in to model training would take 33% off and annual credit subscriptions up to 25%, but neither is part of our scenario. Real-time Melia 1 is $0.40 an hour; Enhanced, with custom dictionary and real-time speaker labels, is $0.75 for batch and $0.80 for real-time. Translation is $0.65 an hour.
Batch transcripts download as JSON, plain text or SRT, but jobs expire after 7 days, so save results as they finish. The Terms say audio files are deleted seven days after the subscription ends and all content is removed when you delete the account. Refunds of unused credits are not stated.
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.
| Limit | As recorded |
|---|---|
| Billing unit | Per second, from credits at hourly rates; credit price falls with packs and subscriptions |
| Free credit | $100 at sign-up, no card, valid 12 months |
| Streaming rate | Melia 1 $0.40/h; Standard $0.45/h; Enhanced $0.80/h (list) |
| Diarization | Included; real-time speaker labels on Enhanced, Standard and Linden, not Melia 1 |
| Languages | 55+ (pricing page); 56 for Melia 1 |
| Training on your audio | Only if you opt in (33% off) |
| Retention | Batch files 7 days, delete sooner by API; real-time audio not stored |
| HIPAA BAA | Sub-BAA through the Trust Center; plan and process not stated |
| EU processing | EU, US and Australia regions for all customers |
| Self-hosted | On-prem containers, virtual appliance and on-device on Enterprise |
| SOC 2 / ISO 27001 | SOC 2 Type II (Trust Center, under NDA); ISO/IEC 27001:2022 |
| Console SSO | SAML/OIDC on Enterprise, $280/month add-on |
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 Speechmatics 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 Speechmatics yet.
Speechmatics lists 6 plans. The cheapest paid plan, Melia 1, pre-recorded (pay-as-you-go list rate), costs $0.24/hr.
Prices as listed, USD, excluding tax; observed 2026-10-16.
Source: speechmatics.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.
