Introducing Whisp API v2
We're excited to announce Whisp API v2, a complete redesign of our developer API focused on real-time streaming transcription. With sub-100ms latency and a simple WebSocket interface, developers can now build voice-first applications that feel truly instantaneous.
What's New in v2
The new API introduces several major improvements:
- WebSocket Streaming: Real-time bidirectional communication replaces the old REST-based approach, enabling true streaming transcription.
- Sub-100ms Latency: From audio input to transcription output, you'll see results in under 100 milliseconds on average.
- Partial Results: Get word-by-word updates as speech is being processed, perfect for live captioning and interactive applications.
- Speaker Diarization: Automatically identify and label different speakers in multi-person conversations.
- Custom Vocabulary: Boost recognition accuracy for domain-specific terms by providing custom word lists.
Getting Started
Connecting to the Whisp API v2 is straightforward. Establish a WebSocket connection, stream audio data, and receive transcription results in real-time. Our SDKs for JavaScript, Python, and Swift handle the connection management and audio encoding for you.
Authentication
API v2 uses API keys for authentication. Keys can be created and managed in the Whisp Developer Dashboard. Each key can be scoped to specific capabilities and rate limits, making it easy to manage access across different applications and environments.
Pricing
API v2 is priced per minute of audio processed. Volume discounts are available for high-usage applications. Enterprise customers can also access dedicated infrastructure for guaranteed performance and custom SLAs.
Migration from v1
Existing v1 customers can migrate to v2 at their own pace. V1 will continue to be supported until December 2026. Our migration guide provides step-by-step instructions for updating your integration.
Developer Resources
Visit our developer portal for complete documentation, code samples, and interactive playground. Our developer relations team is also available in Discord to help with integration questions.