Pulse
Low-latency multilingual speech-to-text API for real-time transcription and recorded audio
About Pulse
Smallest.ai Pulse is a speech-to-text API for real-time audio, voice agents, and streaming transcription. It offers 64ms ultra-low latency and transcription in 38+ languages, with coverage for accents and dialects worldwide. The product provides automatic speaker diarization, real-time sentiment analysis, and language identification for healthcare, legal, finance, media, customer support, and enterprise workloads. Voice processing workflows can be built with the Node and Python SDKs.
Best for
Best use cases
Real-time streaming transcription
Voice agents
Recorded transcription
Global multilingual audio processing
Features
Real-time low latency
Pulse offers 64ms ultra-low latency for real-time audio, voice agents, and streaming transcription.
Multilingual transcription
Supports 38+ languages and accents and dialects worldwide.
Audio understanding
Supports automatic speaker diarization, real-time sentiment analysis, and language identification.
Developer integration
Node and Python SDKs enable voice processing workflows.
Pros
- 64ms ultra-low latency
- Supports 38+ languages plus accents and dialects worldwide
- Automatic speaker diarization, real-time sentiment analysis, and language identification
- Node and Python SDKs available
Cons
- Pulse Pro supports English only.
- Pulse Pro supports pre-recorded audio only.
Pricing Plans
Pulse
- Live streaming & recorded
- 64ms ultra-low latency
- 38+ languages
Pulse Pro
- Pre-recorded audio only
- Maximum transcription accuracy
- Available in English only
Pay as you go
- No monthly commitments
- Scale with your usage
- Access to all features
* Pricing information is for reference only. Please refer to the latest announcement on Pulse's official website.