Google Cloud Speech-to-Text review
by Google LLCSpeech-to-text APIs
Google Cloud's speech-to-text API at $0.016 a minute for batch, sync or streaming on the V2 API, or $0.003 a minute in dynamic batch, billed through a Google Cloud account.
- Free plan
- Card required
- Self-serve
- Usage-based
- Public prices
What you're paying for
- Built-in reach (main value)
- Compliance included
- Worth the price
Speech recognition inside your Google Cloud account, with IAM, audit logs, a BAA and EU processing, and dynamic batch at $0.003 a minute.
For teams already on Google Cloud, Speech-to-Text bills through the same account and uses its IAM roles, Cloud Audit Logs, SSO and compliance reports, and the HIPAA BAA covers it at no extra cost. Google's terms say customer data is not used to train models without permission. Dynamic batch transcribes at $0.003 a minute when results can wait, and Chirp 3 brings diarization and automatic language detection, generally available in the US and EU multi-regions.
- Part of your Google Cloud accountIAM predefined and custom roles, Cloud Audit Logs for admin activity and data access, and SSO through Cloud Identity apply to the API without a separate vendor account.Evidence: IAM, audit logging and Cloud Identity docs (2026-10-16)
- No training without permissionService Specific Terms §18: Google will not use Customer Data to train or fine-tune any AI/ML models without prior permission or instruction; Speech-to-Text is also listed under the HIPAA BAA.Evidence: Service Specific Terms, modified Oct 12, 2026, and HIPAA page (2026-10-16)
- Dynamic batch at $0.003 a minuteThe V2 API's dynamic batch processes audio at a lower level of urgency for $0.003 a minute, against $0.016 for standard recognition.Evidence: Pricing page and batch recognition docs (2026-10-16)
- Best for
- Teams already running on Google Cloud that want transcription under their existing account, IAM and BAA.
- Good to know
- PII redaction is not part of the API, and the 60 free minutes a month apply to the V1 API only.
How Google Cloud Speech-to-Text compares
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 | Google Cloud Speech-to-Text | AssemblyAI | Speechmatics | Amazon Transcribe |
|---|---|---|---|---|
| What you're paying for | 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. | 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 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 | 7.8 | 8.4Best | 8.1 | 7.0 |
| Free option | Free plan | Free trial | Free trial | Free trial |
| Card to start | Card required | No card needed | No card needed | –Not measured |
| Getting access | Self-serve | Self-serve | Self-serve | Self-serve |
| Starter priceCheapest paid plan | $0.003/min | $0.15/hr | $0.24/hr | $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 of audio, no minimum; each channel billed | 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, no minimum; two channels at no extra charge |
| Free credit | 60 minutes a month on the V1 API; none on V2 | $50 at sign-up, no card, does not expire | $100 at sign-up, no card, valid 12 months | 60 minutes a month for 12 months |
| Streaming rate | $0.016/min ($0.96/h), the same as batch | 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) | $0.01/min ($0.60/h) |
| Diarization | No separate price; Chirp 3 diarization in batch only | 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 | Included in batch and streaming, up to 30 speakers |
| Languages | Over 125 (product page) | 99 on Universal-2; 18 on Universal-3.5 Pro | 55+ (pricing page); 56 for Melia 1 | About 90 languages across 113 language rows (our count) |
| Training on your audio | Not without your permission; V1 data logging opt-in gives a lower rate | By default; paid accounts opt out online, free accounts cannot | Only if you opt in (33% off) | By default; opt out with an AWS Organizations policy |
| HIPAA BAA | Covered; accepted in the console per the docs | Signed online at no cost on paid plans | Sub-BAA through the Trust Center; plan and process not stated | Accepted in AWS Artifact |
| EU processing | Chirp 3 GA in the eu multi-region | EU endpoint for batch and streaming, same price as US | EU, US and Australia regions for all customers | Six EU regions for batch, four for streaming |
| Self-hosted | Speech-to-Text On-Prem on GKE, through sales | Streaming only, $20,000 upfront commitment | On-prem containers, virtual appliance and on-device on Enterprise | Not offered |
| SOC 2 / ISO 27001 | SOC 2 Type II and ISO 27001 (Compliance Reports Manager) | SOC 2 Type 2; ISO 27001:2022 | SOC 2 Type II (Trust Center, under NDA); ISO/IEC 27001:2022 | SOC 1, 2 and 3 and ISO 27001 (AWS Artifact) |
| Console SSO | SAML through Cloud Identity Free, no listed cost | SAML/OIDC add-on, $199/month per connection | SAML/OIDC on Enterprise, $280/month add-on | IAM Identity Center, no extra charge |
| Bottom line | 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, 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, 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: Google Cloud Speech-to-Text: 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; 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.
Score
- Pricing clarity
- 7.0 / 10, Good
- batch rate published; streaming rate published; bought online; the V2 price footnote does not name chirp_3; the product page prices Chirp 3 at $0.016/min; custom model training, deployment and inference charges are named but not priced; the product page mentions region-specific pricing; the pricing page shows one price
- Free credit
- 5.5 / 10, Fair
- 60 free minutes a month = 12 hours in the first 12 months; card not stated as optional; no end stated; V1 API only; the V2 API used in the scenario has no free tier
- Price for 500 hours
- 8.0 / 10, Excellent
- $90/mo: Chirp 3 (V2) in dynamic batch at $0.18/h × 500 hours; $0.003/min dynamic batch, flat; standard V2 batch at $0.016/min would be $480; diarization has no separate price; assumes dynamic batch applies to Chirp 3, as the batch docs example shows (the price footnote names "chirp", not chirp_3), and Chirp 3 word-level timestamps work in batch, which its model page lists as unsupported yet describes as available there
- Transcription features
- 8.5 / 10, Excellent
- streaming included; streaming diarization included; custom vocabulary included; PII redaction not offered; language detection included; multichannel included; 125 batch languages ("over 125 languages" on the product page, which also says "85+ languages and variants")
- Data use and retention
- 8.0 / 10, Excellent
- not used for training by default; no delete API stated; retention stated; DPA with the terms; no training without permission (data logging is an opt-in on V1); async transcripts kept about 5 days per the V1 FAQ; V2 retention and a delete endpoint not stated
- Compliance and deployment
- 10.0 / 10, Excellent
- own SOC 2 Type II, ISO 27001, HIPAA BAA online, EU region self-serve, self-hosted option, SAML SSO on a published plan or add-on, BAA accepted in the console per the docs, through account managers per the compliance page; Chirp 3 GA in the eu multi-region; On-Prem through sales; SSO through Cloud Identity Free
Review
Review from public pages
Written from Google Cloud Speech-to-Text'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.
What the public pages show
We read Google Cloud's Speech-to-Text pricing page and product page, the docs on the V2 API, Chirp 3, batch recognition, diarization, multichannel, supported languages, quotas, data usage, data logging, endpoints, custom models and On-Prem, the Google Cloud free-tier page, the Service Specific Terms, the Cloud Data Processing Addendum and the compliance pages 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, and its list prices match the Internet Archive copy of Jul 11, 2026.
The price table is short: one per-minute rate for V2 standard recognition, with volume tiers above 500,000 minutes a month, and a flat dynamic batch rate. No feature is charged as an add-on; the bill depends on minutes, channels, model and batch method.
The model pages take more reading. For en-US, the V2 language table lists speaker diarization only for Chirp 3 and the medical conversation model, and the Chirp 3 page limits diarization to batch. The same page lists word-level timestamps under features it doesn't support, then says they can be enabled in batch. The price footnote names "chirp" but not chirp_3, while the product page prices Chirp 3 at $0.016 a minute.
Strengths
- Low rate when results can wait
Dynamic batch costs $0.003 a minute, so our 500-hour scenario comes to $90 a month (computed), close to the lowest here.
vs Amazon Transcribe: Amazon Transcribe batch is $0.006 a minute ($180) and Deepgram Nova-3 $0.0043 ($129).
- No training without permission
Google does not use customer data to train models unless you opt in; on V1, data logging is an opt-in that lowers the rate.
vs Amazon Transcribe: Amazon Transcribe and Deepgram use customer audio by default unless you opt out.
- Inherits Google Cloud controls
IAM roles, Cloud Audit Logs, customer-managed encryption keys, SSO through Cloud Identity, and SOC 2 and ISO 27001 reports from the Compliance Reports Manager.
- No add-on charges
Diarization, speech adaptation, automatic language detection and multichannel have no separate price; each channel is billed as audio.
Things to know
- No PII redaction in the API
The Speech-to-Text docs do not offer redaction of personal data from transcripts.
Workaround: Plan a separate redaction step after transcription.
- Features vary by model and mode
For en-US, diarization is listed on Chirp 3 (batch only) and the medical conversation model, and Chirp 3 lists word-level timestamps as unsupported while describing them in batch, where files are limited to 20 minutes with them on.
Workaround: The Chirp 3 page says the locations API returns the current features per model and region.
- Free minutes on V1 only
The 60 free minutes a month apply to the V1 API; the V2 API, which carries Chirp 3 and dynamic batch, has no free tier. The $300 Google Cloud welcome credit needs a card and lasts 90 days.
- The standard rate is the highest here
Without dynamic batch, V2 recognition is $0.016 a minute for the first 500,000 minutes, so the same 500 hours cost $480 a month (computed).
Workaround: Dynamic batch, the volume tiers above 500,000 minutes or a Flexible Savings Plan lower it.
Pricing reality
Google bills each second of audio processed, per channel, through your Google Cloud billing account. For our scenario, 500 hours of pre-recorded English with diarization and word timestamps on Chirp 3 in dynamic batch cost 30,000 × $0.003 = $90 a month (computed) before tax, if dynamic batch applies to Chirp 3 as the batch docs show; at the standard V2 rate it would be 30,000 × $0.016 = $480. Streaming costs the same $0.016 a minute. Volume tiers start above 500,000 minutes a month ($0.01, then $0.008 and $0.004). A Flexible Savings Plan takes 10% off for one year or 20% for three and cannot be cancelled. Medical models are $0.078 a minute. Cloud Storage for audio and results is billed separately.
Exit path
Results come back as JSON, inline or written to your own Cloud Storage bucket, and V2 batch can also write SRT and WebVTT captions. Operations can be polled, but listing past transcripts is not described. Under the Cloud Data Processing Addendum, data is deleted within 180 days after the account ends, after a recovery period of up to 30 days.
Compared with
- Pick Google Cloud Speech-to-Text when
- You want audio kept out of training by default and dynamic batch at $0.003 a minute.
- Pick Amazon Transcribe when
- You run on AWS and want PII redaction as a priced add-on and a documented delete API.
- Pick Google Cloud Speech-to-Text when
- You want training off by default and Google Cloud IAM, audit logs and SSO at no listed cost.
- Pick AssemblyAI when
- You want PII redaction, a $50 credit with no card and a BAA signed online in a speech-only account.
- Pick Google Cloud Speech-to-Text when
- You want dynamic batch pricing and a BAA within your Google Cloud account.
- Pick Speechmatics when
- You want a $100 sign-up credit, a stated 7-day batch retention and a delete-job API.
Choose it if
- You already run on Google Cloud and want transcription under the same account, IAM and BAA.
- Your batch jobs can wait and you want the $0.003-a-minute dynamic batch rate.
- You want customer audio kept out of training by default.
Look elsewhere if
- You need PII redaction from the speech API itself.
- You want a free tier on the V2 API.
Bottom line
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.
Plan limits at a glance
| Limit | As recorded |
|---|---|
| Billing unit | Per second of audio, no minimum; each channel billed |
| Free credit | 60 minutes a month on the V1 API; none on V2 |
| Streaming rate | $0.016/min ($0.96/h), the same as batch |
| Diarization | No separate price; Chirp 3 diarization in batch only |
| Languages | Over 125 (product page) |
| Training on your audio | Not without your permission; V1 data logging opt-in gives a lower rate |
| Retention | Sync and streaming not stored; async transcripts about 5 days (V1) |
| HIPAA BAA | Covered; accepted in the console per the docs |
| EU processing | Chirp 3 GA in the eu multi-region |
| Self-hosted | Speech-to-Text On-Prem on GKE, through sales |
| SOC 2 / ISO 27001 | SOC 2 Type II and ISO 27001 (Compliance Reports Manager) |
| Console SSO | SAML through Cloud Identity Free, no listed cost |
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 Google Cloud Speech-to-Text 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 Google Cloud Speech-to-Text yet.
Google Cloud Speech-to-Text lists 6 plans. The cheapest paid plan, V2 dynamic batch, costs $0.003/min.
- Google Cloud Speech-to-TextFreeV1 free tier$0/mo60 minutes a month, V1 API only
- Standard and medical models on V1
- No free tier on the V2 API
- Google Cloud Speech-to-TextV2 standard recognition$0.016/minFirst 500,000 min a month; then $0.01, $0.008 and $0.004
- Batch, sync and streaming at one price
- Chirp 3 priced at $0.016 on the product page
- Diarization and speech adaptation with no separate price
- Google Cloud Speech-to-TextV2 dynamic batch$0.003/minFlat rate; lower urgency, batch recognition only
- Results in Cloud Storage or inline
- SRT and WebVTT caption output
- Google Cloud Speech-to-TextMedical models$0.078/minMedical Conversation and Medical Dictation; first 60 min a month free
- en-US only
- Google Cloud Speech-to-TextFlexible Savings PlanMonthly commitment10% off for 1 year, 20% off for 3 years; cannot be cancelled
- Unused commitment does not carry over
- Overage at on-demand rates
- Google Cloud Speech-to-TextLarge workloadsContact sales
- Additional volume discounts
- Speech-to-Text On-Prem
- Custom Speech-to-Text models (Preview)
Prices as listed, USD, excluding tax; observed 2026-10-16.
Source: cloud.google.com pricing page. Prices change; the observed date above is when our agent read them.