Hermes SDK
AI Noise Cancellation
Removes background noise using advanced AI models while preserving voice integrity without artifacts.
AUDIO PROCESSING MODULES
Modular components you can embed directly into your audio pipeline.
Hermes SDK
HD Voice Boost
Enhances vocal frequencies and balances output levels to deliver clear, consistent voice quality across conditions.
Hermes SDK
Speech Speed Adjustment
Dynamically adjusts speech rate in real time while preserving pitch, prosody, and natural rhythm for better comprehension.
Orpheus SDK
Voice Isolation & Noise Cancellation
Isolates the primary speaker and suppresses background speakers and environmental noise in real time.
Orpheus SDK
Turn Taking
Detects speaker transitions and segments speech turns to enable natural conversational flow in real time.
Orpheus SDK
VAD (Voice Activity Detection)
Detects the presence or absence of speech frame-by-frame to enable triggering, segmentation, and efficient processing.
Modules can be used independently or combined in any order.
SDK Architectures
Two specialized SDKs designed for different real-time voice processing workflows.
HERMES SDK
Optimized for human-to-human communication
Enhances live conversations by improving how speech is delivered and perceived in real time. Controls speed, stabilizes audio, and reduces noise so conversations remain clear, natural, and easy to follow.
- AI Noise Cancellation
- HD Voice Boost
- Speech Speed Adjustment
ORPHEUS SDK
Optimized for human-to-machine communication
Prepares audio for downstream systems by isolating speech, structuring interactions, and improving signal quality. Ensures cleaner, more reliable input into voice AI pipelines and real-time processing systems.
- Voice Isolation & Noise Cancellation
- Turn Taking
- VAD (Voice Activity Detection)
REAL-TIME AUDIO PIPELINE
Process audio streams with low latency, on-device, in real time.
- Frame-level processing
- Sub-200ms end-to-end latency
- Modular chaining of components
- On-device execution
- No audio storage or transmission
BUILT FOR ANY ENVIRONMENT
Integrate anywhere. Works across languages, platforms, and architectures.
Universal Integration
Integrate Hecttor SDK into any voice or audio pipeline with minimal effort.
- Works with any ASR, LLM, or voice system
- Fits into existing pipelines
- SDK-first, built for fast implementation
Cross-Platform Support
Run seamlessly across all major platforms and environments.
- macOS, Windows, Linux
- Mobile and desktop environments
- Browser and native applications
Language & Runtime Agnostic
Use Hecttor SDK with any programming language or runtime.
- Supports Python, C++, Node.js, and more
- Compatible across language versions
- Works with modern and legacy stacks
C++
Python
Node JS
C Sharp
& more...Works with any ASR, LLM or voice stack
PERFORMANCE CHARACTERISTICS
Real-time
Sub-20ms end-to-end latency on live audio streams
On-device
Runs locally within your application environment. No external processing required
No audio storage
Audio is processed in memory. Nothing is stored or transmitted
Robust across inputs
Handles accents, speaking speed, and variable audio conditions
Start building with Hecttor SDKs
Get access to Hermes and Orpheus SDKs and start integrating in minutes.