The Krisp RTC SDK is designed to enhance human-to-human communication by improving audio quality in real-time voice and video applications. It provides CPU-efficient AI models that remove background noise, suppress secondary voices, and even convert accents to create clearer, more inclusive conversations.

Whether you're building a contact-center platform, a conferencing tool, or a telephony application, the Krisp RTC SDK helps you deliver professional-grade audio clarity with minimal engineering overhead.

🚀 What’s Inside the RTC SDK

The RTC SDK includes three core AI models:

Accent Conversion

Transforms an agent’s accent to match the customer’s accent style, improving comprehension and reducing friction in cross-regional communication. Designed for fast, real-time conversion in enterprise telephony and customer-support systems.

👉 Read full docs: Accent Conversion

Noise Cancellation

Removes background noises such as traffic, keyboard clicks, air conditioners, and much more—without distorting the speaker’s original voice. Optimized for CPU execution and low-latency real-time communication.

👉 Read full docs: Noise Cancellation

Background Voice Cancellation (BVC)

Suppresses secondary human voices coming from the environment—such as people speaking near the user—while preserving the primary speaker. This is especially valuable in contact centers, co-working spaces, shared homes, and other noisy environments.

👉 Read full docs: Background Voice Cancellation

🧩 Where the RTC SDK Fits in Your Application

The RTC SDK is designed for on-device and server-side deployment (based on the feature) and supports a wide range of human-to-human communication use cases:

Contact Centers (CCaaS, softphones, PBX systems)

Video conferencing tools

Collaboration platforms

Telephony infrastructure

Real-time streaming and broadcasting

Enterprise communication solutions