kitio.aiAdded Nov 5, 2026
Soniox
PaidMultilingual speech-to-text API for voice agents and transcription
What is Soniox?
Soniox offers real-time and file-based speech recognition across more than 60 languages, with speaker diarization and switching between languages mid-sentence. It is built for developers making voice agents, dictation, meeting transcription and large transcription pipelines.
The v6 model adds agent text input, where a voice agent streams its own replies into the recognizer so it understands both sides of a conversation, and echo cancellation that filters out the agent's voice. It also handles names, numbers and domain terms, and supports real-time translation.
Python and Node SDKs are available, and pricing is token-based, working out to about $0.10 per hour for async files and $0.12 per hour for streaming. Soniox also sells a text-to-speech API and a consumer app.
Features
- Real-time and async transcription
- 60+ languages
- Speaker diarization
- Mid-sentence language switching
- Agent echo cancellation
- Python and Node SDKs
Tags
Releases
+ Submit new release- Listed
Soniox was added to the directory.
Organization
Soniox
Listed by kitio.ai · unverified owner
Pricing
- Pricing model
- Paid
- Paid options from
- Not listed
Soniox plans
Async speech-to-text
$0.10 / hour
- File transcription
- Token-based billing
Real-time speech-to-text
$0.12 / hour
- Streaming transcription
- Same model as async
Text-to-speech
$0.70 / hour
- Generated speech
- Token-based billing
Prices in USD, checked Nov 5, 2026. Plans change often, so confirm on the Soniox site.Official pricing
Sponsored
Your AI tool here
A Featured launch puts your tool in this spot and on the home page for a week.
Top Soniox alternatives
View allAudio & voice
Speech-to-text, text-to-speech and voice agent APIs for developers
Audio & voice
Low-latency text-to-speech API for voice agents
Audio & voice
Low-latency text-to-speech and speech-to-text models for voice apps
Audio & voice
Speech-to-text APIs and voice agents you can run on-premises
If you liked Soniox
Developer platform to build, test and deploy voice AI agents
Freemium · $29/moCustomer support1 viewsMade in United StatesSpeech-to-speech voice agent platform for phone calls
PaidCustomer supportMade in United StatesSpeech-native voice AI platform for real-time voice agents
Freemium · $100/moModels & APIsMade in United States
Pros and cons
Pros
- Very low price per hour
- Strong multilingual support
- Features designed for voice agents
Cons
- Token-based billing takes some math
- Developer API, not a finished app
- Advanced features need integration work
Prompts & results
No prompts yet. Add your own prompt and output to help others use this tool.
Share a prompt
Reviews
Q&A
What is Soniox?
Soniox offers real-time and file-based speech recognition across more than 60 languages, with speaker diarization and switching between languages mid-sentence. It is built for developers making voice agents, dictation, meeting transcription and large transcription pipelines.
Is real-time transcription more accurate than async on Soniox?
No. Both run the same model with the same features; real-time streaming costs about $0.12 per hour and async about $0.10.
Does Soniox offer text-to-speech?
Yes. A token-based TTS API is available, working out to roughly $0.70 per hour of generated speech.
How much does Soniox cost?
Soniox is paid.
Is there a free version of Soniox?
No. Soniox is a paid tool.
What are the main features of Soniox?
Real-time and async transcription, 60+ languages, Speaker diarization, Mid-sentence language switching, Agent echo cancellation, Python and Node SDKs.
Who makes Soniox?
Soniox is made by Soniox.
What are the best alternatives to Soniox?
Popular alternatives include Deepgram, Cartesia Sonic, Cartesia.