Comments
>log in to commentWe started Decibri to simplify Voice AI development by bringing its different technology layers together into one integrated stack. Our goal remains to give developers a unified, neutral foundation that works across programming languages and platforms, with the freedom to choose their own configurations. Users can run models locally, use cloud services, or combine both. Decibri supports its own models, third-party providers, and bring-your-own-model options. We are also making safety and responsible AI a key part of our direction, particularly as Voice AI moves into interactive robotics and physical environments. We would welcome feedback from developers working in this space.
More AI tools →80 free dev tools →
- Real‑time Voice AI and audio processing for Python, Rust, Node.js and browsers
- Cross‑platform microphone capture, streaming and playback
- Built‑in voice activity detection, noise reduction and acoustic echo cancellation
- Vendor‑agnostic integrations with STT/TTS services like OpenAI, AssemblyAI, Deepgram
Decibri provides a cross-platform audio stack for real-time voice AI, handling capture, playback, VAD, echo cancellation and streaming.
- for
- Developers building voice assistants, agents, robots and other real-time voice AI applications.
- pricing
- open source
- license
- Apache-2.0
- npm-v5.8.0a month ago
- crate-v6.4.0a month ago
- npm-v5.7.03 months ago
Key features
- Unified Real-Time Voice AI Stack — Brings audio capture, processing, speech detection, and AI integrations together in one technology stack for building Voice AI applications and intelligent agents.
- Vendor-Neutral Speech AI Integrations — Connect to supported local and cloud-based speech-to-text engines, local text-to-speech models, or custom speech pipelines. Choose providers and models based on your requirements.
- Cross-Platform Development — Build applications using Python, Node.js, and Rust across supported Windows, macOS, and Linux systems. Browser support provides additional audio capture and playback capabilities.
- Low-Latency Audio Streaming and Full-Duplex Communication — Stream audio to AI agents, speech engines, and other services in real time. Capture microphone input and play audio simultaneously for interactive voice applications.
- Advanced Audio Conditioning and Noise Reduction — Improve live and recorded audio with denoising, high-pass filtering, automatic gain control, DC removal, and limiting through Decibri's Audio Conditioning Engine.
- Acoustic Echo Cancellation (AEC) — Reduce speaker audio feeding back into the microphone. Support natural two-way voice interactions where applications need to listen while speaking.
- Multichannel Audio Processing and Resampling — Capture and process multiple audio channels with configurable sample rates and channel selection. Adapt audio formats to different devices and speech processing requirements.
- Audio File Processing — Process recorded audio using the same conditioning capabilities available for live microphone streams. Apply noise reduction, analyze speech activity, and save processed recordings
Use cases
- Real-time conversational AI agents with natural two-way voice interactions
- Interactive robotics and Physical AI systems that listen and respond
- Voice-enabled assistants for desktop, mobile, and web applications
- Fully offline, cloud-connected, and hybrid Voice AI applications
- Real-time speech recognition and transcription pipelines
Decibri vs alternatives
ElevenLabs | ||||
|---|---|---|---|---|
| Best for | Real-time audio I/O for AI agents | Open-source speech recognition | Local neural TTS | AI text-to-speech |
| Pricing | Open source | Free | Free | Subscription |
| DevHunt upvotes | 1 | 0 | 0 | 0 |
| Launched | Nov 2026 | — | — | — |
- Decibri vs OpenAI Whisper: Whisper is a speech-to-text model; Decibri provides the audio capture and processing layer.
- Decibri vs Piper TTS: Piper TTS generates speech; Decibri focuses on audio input and VAD, not synthesis.
- Decibri vs ElevenLabs: ElevenLabs offers cloud TTS and voice cloning, while Decibri handles real-time audio I/O locally.
Decibri FAQ
What is Decibri, and who is it for?+
Decibri is a unified technology stack for developers building real-time Voice AI applications, intelligent agents, and interactive robots. It brings audio capture, processing, streaming, and speech AI integrations together in one cross-platform solution.
How is Decibri different from other Voice AI tools?+
Decibri provides an integrated foundation across the audio and speech processing layers of Voice AI. Unlike standalone speech recognition or text-to-speech tools, it combines real-time audio infrastructure with integrations for different AI engines. Its vendor-neutral approach gives developers greater flexibility over their technology choices.
Can Decibri run completely offline?+
Yes. Decibri supports local audio processing, voice activity detection, speech recognition, text-to-speech, and keyword spotting without cloud services. Developers can build offline, cloud-connected, or hybrid speech pipelines using supported models and providers. Fully offline intelligent agents also require local models for their other AI components.
Which programming languages and platforms does Decibri support?+
Decibri supports Python, Node.js, and Rust across supported Windows, macOS, and Linux systems. Browser support is also available for audio capture, playback, and basic voice activity detection. Some advanced audio processing capabilities are available only through the native implementations.
Can Decibri be used for robotics and Physical AI?+
Yes. Decibri can provide the audio and speech infrastructure for interactive robots and Physical AI systems. Developers can connect these capabilities to their own AI models and control systems to create applications that listen, understand, speak, and act.
How do I get started with Decibri?+
Install Decibri using pip for Python, npm for Node.js, or the Rust package ecosystem. The documentation includes installation instructions, API references, and examples for audio capture, processing, and speech AI integrations. Documentation: https://decibri.com/docs/
Edited by the makers, based on decibri.com · Nov 16, 2026. Details may change; check the official site.
ElevenLabs




