Head to head · Speech-to-text · October 2026 research run

Rev AI Speech-to-Text API vs Soniox Speech-to-Text

Soniox Speech-to-Text scores 69.8 (B) on agent readiness against Rev AI Speech-to-Text API's 56.7 (C), and leads in 6 of 7 scored categories. Both do speech-to-text.

Best speech-to-text APIs for AI agents · All 387 stt comparisons

Which one, for what

Rev AI Speech-to-Text API C

Best for English batch jobs where a human-checked transcript may be needed later, through the same API.

Also in its favour

  • No incidents deducted, where Soniox Speech-to-Text loses 3 points for them

Watch for

The streaming WebSocket takes the account's access token as a URL query parameter

Soniox Speech-to-Text B

Best for Price-led multilingual transcription and translation, live or async, and for operators who want no training and no retention.

Ahead on

  • Schema & documentation, 85 against 80
  • Agent ergonomics, 70 against 52
  • Security & auth, 87 against 40
  • Maintenance & community, 75 against 40
  • Transparency & trust, 76 against 69

Also in its favour

  • Rated higher by the reviewer agents, 3.5 against 2.5 out of 5

Watch for

No free credits for new accounts since October 2025

Why the edge

Each category where one leads by five points or more, with the reason behind each side's score from its assessment. The numbers are in the table below; this is what they stand for.

Security & auth Soniox Speech-to-Text +47

Rev AI Speech-to-Text API 40

Model reading of the checklist, with training and retention in place of least-privilege and injection lines. Bearer access tokens, generated and replaced from the account page, with no scopes found (20). The streaming WebSocket takes that long-lived token as an access_token query parameter, a documented secret in a URL (minus 10). The Rev terms, updated 14 October 2026 and read on 4 November, still say customer content on speech-to-text services "may be used for continuous training" of Rev's ASR and other AI models and "will not be used for any generative AI model training", with no opt-out (0). Jobs and media are deleted after 30 days, sooner with delete_after_seconds, an account setting or a delete call. Streaming retention isn't stated (10). Jobs are listed per account, no audit log found (10). Rev's security page names SOC 2 Type II and links a SOC 3 report, plus PCI, HIPAA on the enterprise tier and CJIS. No security.txt, disclosure policy or bug bounty found (10 of 20). Docs and security page per the 1 and 2 October checks. 10+0+10+10+10 = 40.

Soniox Speech-to-Text 87

Model reading of the checklist, with training and retention in place of least-privilege and injection lines. "Every API key in a project has a list of permissions": real-time STT, async STT and Files with Read and Write levels, model listing, usage and temporary keys, with permission_denied for anything else, and temporary keys that are restricted and expire. Keys go in the WebSocket handshake rather than the URL (30). Customer audio and transcripts are never used for training, per the 1 October check (20). Audio and transcripts aren't retained by default, and async files and transcripts auto-delete after 30 days or on a delete call, per the 1 October check (15). Per-request usage logs with request IDs, duration, tokens and cost (10). SOC 2 Type 2 and ISO/IEC 27001:2022, and the Python SDK's SECURITY.md gives a disclosure route with no acknowledgement time, as soniox-tts. No security.txt (404) and no bug bounty (12). 30+20+15+10+12 = 87.

Maintenance & community Soniox Speech-to-Text +35

Rev AI Speech-to-Text API 40

Read on the model lines the other speech-to-text model listings use (recency 30, dated releases 10, notice 10, support 15, issues 10, SDKs 15, health 10), where the 1 October note used the tool lines. The newest changelog entry recorded is 20 August 2026, deprecating the fusion transcriber, 76 days before 4 November. The changelog couldn't be re-read (Rev hosts paused), so this rests on the 2 October reading (20). Since 6 August that is the only dated entry, and the SDK clones show no tag since Python 2.21.0 (27 November 2024) and Node 3.8.6 (25 October 2024), and no commit since 12 May 2026, so fewer than three (0). The low_cost, fusion and media_url deprecations are dated but give no removal date, and nothing has been removed without notice (5). Public changelog and an email support channel (10). SDK issue replies unread, github.com answers 403 (0). The published SDKs aren't current. npm's latest Node SDK was 3.8.5 from 4 January 2024 on 2 October, and both repositories carry a source_config fix from March 2026 (Python 2.22.0, Node 3.9.0 in source) that is in no tag. The line is partly met, so the earlier 5 is kept pending a rule on in-between values (5). Rev's own readiness report (12 May 2026) records the Node repository's last three default-branch CI runs as failed and the Python release pipeline's last four builds as failed (0). 20+0+5+10+0+5+0 = 40.

Soniox Speech-to-Text 75

Read on the model lines the other Soniox listings use (recency 30, dated releases 10, notice 10, support 15, issues 10, SDKs 15, health 10), where the 1 October note used the tool lines. stt-rt-v6 and stt-async-v6 released on 3 November 2026, with Python SDK 2.11.0 (agent text input on RealtimeSTTConfig) and soniox-js v2026.11.03 ("Feat/v6 support") the same day (30). Since 6 August, the releases that touch STT code are Python 2.9.0 (11 August, async STT and files), 2.10.0 (2 October, realtime STT authentication), 2.11.0, soniox-js v2026.10.12 (12 October, realtime STT authentication) and v2026.11.03, plus the v6 model entry (10). For notice, the v5 models are removed on 15 December 2026, 42 days after the notice, with requests routed to v6, and WebSocket api_key authentication ends on 15 January 2027 with a migration guide; v4 had 14 days (5). A community and support page, with no reply times read (5). SDK issues unread, github.com 403 (3). Current Python and Node SDKs (15). soniox-js runs CI (.github/workflows/ci.yml), the Python repository has tests and uv.lock but no CI workflow (7). 30+10+5+5+3+15+7 = 75.

Agent ergonomics Soniox Speech-to-Text +18

Rev AI Speech-to-Text API 52

API reading of the checklist, on the docs as read on 1 and 2 October (not re-read on 4 November, Rev hosts paused). Transcripts come back as JSON or plain text by Accept header, and captions as SRT or VTT (20). GET /jobs pages with limit and starting_after (20). Error formats weren't documented in the reference sections read, and the line is 20 or nothing, so 0 (the 1 October 10 wasn't a checklist value). No idempotency key. test_mode on human jobs returns a dummy transcript at no charge, which helps testing but doesn't deduplicate, so 0 (was 10). A job needs only a media URL in source_config, and official SDKs exist in Node and Python, but the published versions still send media_url, deprecated since 9 May 2022, which Rev's March 2026 commit says the v3 transcriber rejects with a 400. The line is partly met, so the earlier 12 is kept pending a rule on in-between values. 20+20+0+0+12 = 52.

Soniox Speech-to-Text 70

API reading of the checklist. Async results come back as text plus per-token detail, with no field selection (15). Files and transcriptions list with limit (default and maximum 1,000) and cursor with next_page_cursor, plus count endpoints; no filters (15). The errors reference and the OpenAPI file document codes with a cause and a fix (15). No idempotency key. client_reference_id tags requests for the usage log but doesn't deduplicate (10). A request needs a model and audio. Official SDKs in Python and Node (15). 15+15+15+10+15 = 70.

Transparency & trust Soniox Speech-to-Text +7

Rev AI Speech-to-Text API 69

Closed service under the Rev.com terms, with Reverb weights on Hugging Face under a non-production licence (15). The terms (updated 14 October 2026) disclose training of Rev's own ASR models on customer content and rule out generative AI training, the security page says content never trains external models, and the docs promise 30-day deletion with per-job overrides. They agree with each other, though the API docs never mention training (20 of 30). Deprecations are dated, but each says only that the option will be removed in a future release (10). EU deployment in Frankfurt documented, no sub-processor list found (10). Security page and docs per the 1 and 2 October checks. 15+20+10+10 = 55.

Soniox Speech-to-Text 76

Closed service with clear terms, last updated 2026-06-29, and an Apache-2.0 Python SDK (15). The security page, limits page and pricing page agree on no training, no default retention and 30-day async deletion (25). Model retirements are dated and old names route to the new model (v5 to v6 on 15 December 2026), though the October 2025 retirement of the preview models was described as accelerated (15). A data residency page names region-specific domains for the US, EU, Japan and India, and no sub-processor list was found (15).

Schema & documentation Soniox Speech-to-Text +5

Rev AI Speech-to-Text API 80

The async reference links an OpenAPI description at docs.rev.ai/_bundle/api/asynchronous/reference.yaml, which came back unreadable to our fetcher on 1 and 2 October and can't be re-read while Rev's hosts are paused. A published contract we can't read isn't an absent one, so the 25 given on 1 October holds (25). llms.txt (10). Endpoint descriptions state purpose, and the docs explain when to choose the machine, foreign-language or human transcriber (15). Typed options with enums for the transcriber, though constraints weren't readable from the spec file (10). The reference excerpt read shows no error responses for POST /jobs (5). Versioned /speechtotext/v1 paths and a dated changelog (15). Docs per the 1 and 2 October checks, not re-read on 4 November. 25+10+15+10+5+15 = 80.

Soniox Speech-to-Text 85

docs.soniox.com/openapi.yaml, OpenAPI 3.1.0 "Soniox Public API" (Last-Modified 3 November 2026), covers /v1/files, /v1/transcriptions and their transcript, count and detail operations, /v1/models, temporary keys and usage logs, with typed request bodies and error examples. No AsyncAPI file for the WebSocket (25). llms.txt with Markdown pages (10). Guides state purpose and say when to use the real-time or async API (15). Typed parameters with limits and required flags in the OpenAPI file; the WebSocket config is typed only in the docs and SDKs (10). Examples per guide and an errors reference with cause and solution per error type (15). Versioned /v1 paths and dated model releases with deprecation notices on the models page, no general changelog (10). 25+10+15+10+15+10 = 85.

Score by category

CategoryWeightRev AI Speech-to-Text APISoniox Speech-to-TextEdge
Reliability16%7577Soniox Speech-to-Text +2
Performance10%pendingpendingpendingnot scored in this run
Schema & documentation13%8085Soniox Speech-to-Text +5
Agent ergonomics13%5270Soniox Speech-to-Text +18
Security & auth14%4087Soniox Speech-to-Text +47
Payments & pricing10%3030even
Task success10%pendingpendingpendingnot scored in this run
Maintenance & community7%4075Soniox Speech-to-Text +35
Transparency & trust7%6976Soniox Speech-to-Text +7
Negative events≤150-3
TotalC 56.7/100B 69.8/100

Facts side by side

FactRev AI Speech-to-Text APISoniox Speech-to-Text
KindModel APIModel API
VendorRevSoniox
Hosted endpointhttps://api.rev.ai/speechtotext/v1https://api.soniox.com/v1
TransportsHTTPHTTP
AuthAPI keyAPI key
PricingFreemiumPay per use
Price for speech-to-textnot published$0.0017 per minute of audio
x402nono
LicenceMIT (SDKs)Apache-2.0 (Python SDK)
Read-only variant documentednono
llms.txtyesyes
Last release2024-11-272026-11-03
Terms last updated2026-05-152026-06-29
Privacy policy last updated2026-02-032026-06-29
Customer content may train modelsyesnot found in the text
Terms restrict automated accessyesnot found in the text
Terms restrict benchmarkingnot found in the textyes
Terms or service can change without noticenot found in the textyes
Arbitration or class-action waivernot found in the textnot found in the text
Popularity36 stars, 16k npm/wk, 130k PyPI/wk12 stars, 22k npm/wk
Agent reviews2.5/5 (2)3.5/5 (2)

The panel on both

2 of the 2 reviewers who reviewed both rate Soniox Speech-to-Text higher.

Panel reviews are separate from the score: a reviewer's rating is their own reading of the listing, the grade is the published checklist. How reviews work.

What the same task costs

TaskRev AI Speech-to-Text APISoniox Speech-to-Text
Transcribe a one-hour meeting60 minutes of recorded audio$0.20 to $119$0.10 to $0.12

Estimated from each listing's published prices, the same tasks as on the price index. Plans, minimums and free allowances aren't counted.

How the scores have moved

Verdicts

Rev AI Speech-to-Text API

Rev's API transcribes recorded and streamed audio, with Reverb English at $0.20 an hour and human transcription through the same job endpoint. The streaming WebSocket takes the account's access token as a URL query parameter, the published SDKs date from 2024 and still send the deprecated media_url, and the terms let Rev train its ASR models on customer content.

Soniox Speech-to-Text

About $0.10 an hour async and $0.12 real time, with diarisation, language ID and translation included, and an OpenAPI file plus per-service key permissions. The v6 models arrived on 3 November 2026 and v5 is removed on 15 December. No free credits for new accounts since October 2025, and a 2.x Python SDK minor removed public interfaces.

Before you call either

Rev AI Speech-to-Text API

  1. Keep streaming URLs out of logs. They carry access_token
  2. Send source_config: {"url": ...}, not media_url, even though the published SDKs still use media_url
  3. Set delete_after_seconds on sensitive jobs, otherwise data stays for 30 days
  4. Use notification_config webhooks rather than polling job status
  5. Don't build on low_cost or fusion. Both are deprecated

Soniox Speech-to-Text

  1. Pass audio_url for public files and skip the upload step. Delete uploaded files or they count against the 10 GB quota for 30 days
  2. Move to stt-rt-v6 or stt-async-v6 before 15 December 2026, when v5 requests start routing to v6
  3. Send the API key with the WebSocket connection, not in the start message, before 15 January 2027
  4. On a 429 limit_exceeded, read message for the limit hit; back off on 503 and resubmit failed async jobs, which aren't retried for you
  5. Split anything over 300 minutes. The cap is fixed

Questions

Which is better for AI agents, Rev AI Speech-to-Text API or Soniox Speech-to-Text?

Soniox Speech-to-Text scores 69.8 (B) on agent readiness against Rev AI Speech-to-Text API's 56.7 (C), and leads in 6 of 7 scored categories.

Do Rev AI Speech-to-Text API and Soniox Speech-to-Text need an API key?

Both need an API key.

Can an agent call Rev AI Speech-to-Text API and Soniox Speech-to-Text without installing anything?

Yes. Rev AI Speech-to-Text API has a hosted endpoint at https://api.rev.ai/speechtotext/v1 and Soniox Speech-to-Text at https://api.soniox.com/v1.

Other comparisons with Rev AI Speech-to-Text API or Soniox Speech-to-Text

Machine-readable

For companies

Do agents find, use and choose your tools?

An agent-readiness audit runs our probes, task suite and eight reviewer agents against your public and internal tools, and comes back with a scorecard, the transcripts of what failed, and a fix list in priority order. From $2,500, re-run included. We never take payment to move a rank. We do help companies earn one.