Skip to main content
ToolPotion
AI Advisor
Submit a Tool (Free)

Decibri

AI AppsAI Agent BuildersAudio, Voice & Music#45Free · Windows, macOS, Linux, Browser

Decibri is a unified audio layer for AI agents and Voice AI applications, offering real-time microphone capture, speaker output, and voice activity detection. It provides cross-platform support for Python, Node.js, and Rust, with pre-built binaries for easy installation. Designed for real-time systems, Decibri simplifies audio infrastructure for voice applications, enabling seamless integration with various speech and audio processing engines.

Visit Website
Share
Decibri screenshot
Decibri screenshot

Decibri serves as a unified audio layer specifically designed for AI agents and Voice AI applications. It provides essential real-time audio capabilities, including microphone capture, speaker output, and voice activity detection (VAD). The platform supports multiple programming languages, namely Python, Node.js, and Rust, making it versatile for various development environments.

Read More

One of Decibri's key advantages is its ease of installation and cross-platform compatibility. It offers pre-built binaries for Python and Node.js across all supported platforms (Windows, macOS, Linux), eliminating the need for compilation or complex build tools. This "zero setup" approach ensures developers can quickly integrate real-time audio functionalities into their projects. The Rust core, which powers the Python, Node.js, and Rust packages, compiles to native code, ensuring predictable low-latency audio directly from the operating system's audio layer.

Decibri captures audio directly from the OS audio layer, providing 100ms audio chunks by default in 16-bit signed integer PCM format, which is the standard expected by most speech and wake-word engines like Vosk, sherpa-onnx, and whisper.cpp. A 32-bit float format is also available. It supports stream-native idioms for each language, allowing developers to pipe audio to files, WebSockets, or voice agents efficiently. The platform also includes a bundled Silero VAD v6.2 model for accurate speech detection in noisy environments, running locally via ONNX Runtime without requiring cloud APIs.

Beyond basic capture and playback, Decibri offers advanced audio processing features through its Audio Capture Engine (ACE), including DC removal, denoise, high-pass filtering, automatic gain control, and limiting. It also provides Acoustic Echo Cancellation (AEC) to remove loudspeaker sound from microphone capture. For offline processing, the Audio File Processor (AFP) allows conditioning recorded or in-memory samples. Decibri integrates with numerous speech-to-text (STT) providers, both cloud-based (AssemblyAI, AWS Transcribe, Azure AI Speech, Deepgram, Google Cloud Speech-to-Text, Mistral Voxtral, OpenAI Realtime API) and local (Sherpa-ONNX, Whisper.cpp), as well as local text-to-speech (TTS) with the Kokoro model and keyword spotting (KWS) with Sherpa-ONNX.

Decibri is built for real-time systems, ensuring low-latency audio I/O without external audio tools or platform-specific handling. It supports asynchronous operations in Python with `asyncio` and works in both Node.js and browser environments from a single package. The `decibri-cli` tool provides command-line capabilities for audio capture, playback, and device enumeration, further extending its utility for scripting and CI/CD pipelines.

  • CategoryAI Agent Builders
  • Pricing ModelFree
  • DeploymentLocal
  • PlatformsWindows, macOS, Linux, Browser

  • Real-time Microphone Capture

  • Speaker Output and Playback

  • Voice Activity Detection (VAD)

  • Cross-platform Support (Python, Node.js, Rust)

  • Pre-built Binaries for Zero Setup

  • Direct OS Audio Layer Access

  • 16-bit PCM and 32-bit Float Audio Output

  • Asynchronous API Support (asyncio for Python)

  • Browser Compatibility (Node.js package)

  • Audio Capture Engine (DC removal, denoise, AGC)

  • Acoustic Echo Cancellation (AEC)

  • Local ML Voice Detection (Silero VAD)

  • Integration with Cloud STT Services

  • Integration with Local STT/TTS/KWS Models

  • Command-line Interface (decibri-cli)

  1. Install Decibri: Use `pip install decibri` for Python or `npm install decibri` for Node.js to get the pre-built binaries.

  2. Open Microphone: Utilize `decibri.Microphone.open()` in Python or `new Microphone()` in Node.js to start capturing audio.

  3. Stream Audio: Read audio chunks using Python iterators, Node.js Readable streams, or Rust channels.

  4. Process Audio: Apply built-in conditioning like DC removal, denoise, or acoustic echo cancellation.

  5. Detect Speech: Enable the bundled Silero VAD model for local voice activity detection.

  6. Integrate with AI Services: Pipe raw PCM audio to speech-to-text services, voice agents, or other AI models.

  7. Play Audio: Use Decibri's Speaker API to play audio through the system speaker.

  8. Record to File: Employ `record_to_file()` for simple audio recording or full streaming control for advanced use cases.

  • Real-time Voice AI Agents
  • Speech-to-Text Integration
  • Voice Activity Detection
  • Audio Recording for ASR
  • Cross-platform Audio Applications
  • Server-side Voice Backends
  • Keyword Spotting
  • Full Duplex Audio Communication

Decibri Reviews

Loading...

Popular AI Tools Like Decibri

Deepgram Voice Agent API enables the creation of real-time voice AI agents using advanced speech-to-text and text-to-speech technologies. It offers seamless orchestration for…

AI Voice Agents
View Tool

Enables the creation of real-time voice AI agents using advanced speech-to-text and text-to-speech technologies. It offers seamless orchestration for…Deepgram Voice Agent API enables the creation of real-time voice AI agents using advanced speech-to-text and text-to-speech technologies. It offers seamless orchestration for…

AI Voice Agents
See the Top 250 AI Apps

Open Voice OS is a community-driven, open-source voice AI platform for building custom voice-controlled interfaces. It offers NLP, a customizable UI, and prioritizes privacy and…

AI Voice Agents

Ranked #89 in AI Voice Agents

View Tool

Community-driven, open-source voice AI platform for building custom voice-controlled interfaces. It offers NLP, a customizable UI, and prioritizes privacy and…Open Voice OS is a community-driven, open-source voice AI platform for building custom voice-controlled interfaces. It offers NLP, a customizable UI, and prioritizes privacy and…

AI Voice Agents
Ranked #89 in AI Voice Agents

Voice.ai

AI Apps

Voice.ai is a real-time AI voice changer and voice platform offering live voice transformation across 50+ apps, free voice cloning, text-to-speech, voice agents, and developer…

AI Audio EditorsGaming & Esports

Ranked #239 in AI Apps

View Tool

AI Apps

Real-time AI voice changer and voice platform offering live voice transformation across 50+ apps, free voice cloning, text-to-speech, voice agents, and developer…Voice.ai is a real-time AI voice changer and voice platform offering live voice transformation across 50+ apps, free voice cloning, text-to-speech, voice agents, and developer…

AI Audio EditorsGaming & Esports
Ranked #239 in AI Apps

Aethex

AI Apps

AethexAI provides end-to-end voice agents tailored for emerging markets. With proprietary models and localized solutions, it enables businesses to deploy voice technology…

AI Voice AgentsFinance & Banking

Ranked #167 in AI Voice Agents

View Tool

AI Apps

AethexAI provides end-to-end voice agents tailored for emerging markets. With proprietary models and localized solutions, it enables businesses to deploy voice technology…

AI Voice AgentsFinance & Banking
Ranked #167 in AI Voice Agents

Resum

AI Apps

Resum provides an enterprise API and SDK designed to connect live conversations with named AI agents. It captures audio from various sources, generates a real-time transcript, and…

AI Voice Agents
View Tool

AI Apps

Provides an enterprise API and SDK designed to connect live conversations with named AI agents. It captures audio from various sources, generates a real-time transcript, and…Resum provides an enterprise API and SDK designed to connect live conversations with named AI agents. It captures audio from various sources, generates a real-time transcript, and…

AI Voice Agents
See the Top 250 AI Apps

VideoSDK

AI Apps

A developer platform with APIs and SDKs for real-time voice, video, and AI agents, letting teams embed audio/video calling, live streaming, and AI voice agents into web and mobile…

AI DevOps & Cloud Tools
View Tool

AI Apps

A developer platform with APIs and SDKs for real-time voice, video, and AI agents, letting teams embed audio/video calling, live streaming, and AI voice agents into web and mobile…

AI DevOps & Cloud Tools
See the Top 250 AI Apps

Moonshine offers very low latency speech-to-text, intent recognition, and text-to-speech capabilities, enabling the creation of efficient voice agents and interfaces. Ideal for…

AI Voice Agents
View Tool

Offers very low latency speech-to-text, intent recognition, and text-to-speech capabilities, enabling the creation of efficient voice agents and interfaces. Ideal for…Moonshine offers very low latency speech-to-text, intent recognition, and text-to-speech capabilities, enabling the creation of efficient voice agents and interfaces. Ideal for…

AI Voice Agents
See the Top 250 AI GitHub Repos

iSpeech offers a powerful Text to Speech API and Speech Recognition SDK, enabling natural-sounding voice synthesis and speech recognition in over 30 languages. Ideal for…

Text to Speech
View Tool

Offers a powerful Text to Speech API and Speech Recognition SDK, enabling natural-sounding voice synthesis and speech recognition in over 30 languages. Ideal for…iSpeech offers a powerful Text to Speech API and Speech Recognition SDK, enabling natural-sounding voice synthesis and speech recognition in over 30 languages. Ideal for…

Text to Speech
See the Top 250 AI Apps
1 of 8