Who Soniox is best for
Soniox sells speech recognition, translation and text-to-speech through an API, plus the Soniox App for individuals and teams. Medical transcription is one of its listed use cases, but it is not a clinical documentation product: it does not write notes or connect to an EHR. Healthcare use usually comes through vendors that build dictation or scribe tools on the API.
Best for
- Health tech companies building dictation, transcription or scribe products that need a speech-to-text engine
- Developers who need real-time transcription across many languages, including mid-conversation language switches
- Multilingual clinicians who want inexpensive voice typing
Not ideal for
- Practices that want a ready-made ambient scribe that writes notes into the EHR
- Organizations that need a signed BAA confirmed before evaluating a tool
Soniox pros and cons
Pros
- Published pay-as-you-go API pricing: $0.10 per audio hour for files and $0.12 for real-time
- 60+ languages in one model, with speaker separation and language identification included
- Per-session context for medication and procedure names
- Lists SOC 2 Type 2, ISO/IEC 27001:2022, HIPAA and GDPR compliance
- Says customer audio is never used to train its models
- App has a free plan and runs on iOS, Android, macOS, Windows and the web
Cons
- Not a medical product: no note templates, coding support or EHR integration
- Its public pages do not say whether Soniox signs a business associate agreement
- The medical page publishes no accuracy figures beyond a customer testimonial
Soniox pricing
API (pay-as-you-go)
- Speech-to-text, async (files): $0.10 per audio hour
- Speech-to-text, real-time: $0.12 per audio hour
- Diarization, language identification and formatting included; advanced use is billed by token
- Text-to-speech: about $0.70 per hour of generated speech
- Enterprise: quote-based
Soniox App
- Free: $0, with Smart Scribe, Translator and Voice Typing on weekly free credits
- Pro: $19.99 per month for unlimited use, priority processing and early access to features
- Business: $30 per user per month (billed monthly), adding seat management, shared workspaces and data residency options
Key features
- Speech-to-Text API: real-time and file transcription in 60+ languages
- Context: per-session vocabulary such as medication and procedure names
- Speech Translation: real-time translation across 3,600 language pairs, according to the vendor
- Text-to-Speech: streaming speech with voice cloning and more than 200 built-in voices
- Data residency: in-region processing in the US, EU, Japan and India
- Soniox App: Smart Scribe (conversation capture), Translator and Voice Typing, which types into any text field on desktop or mobile
HIPAA, security and integrations
Soniox's security documentation states that it meets SOC 2 Type 2, ISO/IEC 27001:2022, HIPAA and GDPR. It says real-time audio is processed transiently, async files are deleted after 30 days and customer data is never used to train models. Its medical page says healthcare use may require a business associate agreement, but no page we reviewed confirms Soniox signs one. Ask sales before sending PHI.
Soniox has no native EHR integrations; your application maps API transcripts into EHR or scribe systems. The app's Voice Typing works system-wide, so it can type into an EHR text field.
Soniox alternatives
- Dragon Medical One — cloud medical dictation with clinical vocabularies, sold to clinicians rather than developers.
- Microsoft Dragon Copilot — dictation plus ambient note generation with EHR integration.
- Augnito — healthcare-specific speech recognition with an SDK and a clinician app.
- Freed — ready-made AI scribe for small practices; no development needed.
