whereismyai
Human or AI?
Human or AI? is an audio analysis tool that identifies whether a spoken voice belongs to a human or an artificial intelligence.
Voice Check listens to playing audio and determines whether a voice is human or artificial as it speaks. It runs locally in the browser or via a macOS app called Off the Record, ensuring audio stays on the user device without leaving recordings. The interface displays the last minute of audio readings color-coded in green for human and red for AI. It works across languages by analyzing voice sound rather than content. Developers can also integrate the underlying models via API or local server deployment.
Human or AI? at a glance
- Best for: Software engineer
- Runs on: Web, macOS, API, self-hosted
- Pricing: 1,000 free API credits to start
- Released: October 2026
- Category: AI chat assistant
- Website: weirdmachines.ai/voice-check
Key features
- Listens to browser tab audio in real time
- Displays a green and red visual reading of the last minute
- Runs locally in the browser without sending audio away
- Provides a macOS menu bar integration
- Analyzes voice characteristics independently of language
- Offers a developer API for voice detection
- Supports local deployment on custom servers
Use it for
- Detecting AI-generated voices during live calls and videos
- Checking audio content in browser tabs
- Monitoring voice messages on macOS
- Integrating voice detection into third-party apps via API
Alternatives to Human or AI?
- AssemblyAI — Industry-leading speech models deliver high accuracy and low latency for voice applications.
- FluidAudio — Runs local audio AI models on Apple Neural Engine.
- Pipecat — Builds real-time voice and multimodal agents with frame-level control.
- NovaSynth — Tests voice agents on real calls and validates fixes automatically.
- Deep Chat Component — Connects directly to any API and popular AI providers from the browser.
- Chorus — Compare multiple AI chat outputs side by side without switching browser tabs.
- WeNet — Achieves state-of-the-art results on public speech datasets.
- Auricle — Runs entirely in the browser via Rust and WebAssembly with low latency and high accuracy.
- Coval — Prove voice agent reliability through simulation and human-in-the-loop testing.
- m’kay — Talk to your Mac's AI agents by voice from anywhere.
- Chatbox — Chat with multiple AI models and local documents using your own API keys.
- sherpa-onnx — Runs complex audio AI tasks offline on local devices and embedded hardware.
All Human or AI? alternatives