A cross-platform, long-running microphone recorder, Eve addresses the need for persistent, searchable voice data workflows. It targets users like researchers, students, and professionals requiring easily manageable and searchable voice logs for meetings, interviews, or personal notes. Eve enables effortless preservation and retrieval of voice data, with the flexibility to re-transcribe using future, more advanced AI models.

How It Works

Eve employs continuous recording, segmenting audio into FLAC files. It uses Voice Activity Detection (VAD) with Silero VAD to isolate speech segments, skipping silence and reducing noise. Real-time transcription is handled by Qwen3-ASR by default, processing speech-only chunks. An automatic microphone switching feature selects the active input source. Audio and transcription files are archived chronologically, with optional OneDrive synchronization.

Quick Start & Requirements

• Install: Requires Python >= 3.12 and uv. Install uv (e.g., brew install uv), then run uv sync from the project root.
• Run: uv run eve
• Prerequisites: Microphone permissions, network for ASR model download. GPU/NPU recommended for ASR performance.
• Resource Guidance: Recording-only: ~2GB RAM/dual-core CPU. Real-time ASR (CPU): ~8GB RAM/4+ cores. GPU/NPU ASR: ~8GB RAM. Disk space varies; 10GB+ recommended.
• Links: No direct quick-start/demo links. Installer build scripts available.

Highlighted Details

• OneDrive Sync: Output directory can be set to a local OneDrive folder for automatic cloud sync.
• Offline Transcription: Supports asynchronous transcription via eve transcribe.
• Cross-Platform Installers: Scripts and CI build native installers (.pkg, .deb, .exe).
• Automatic Microphone Switching: Intelligently detects and switches to the active microphone input.
• VAD-based Speech Segmentation: Transcribes only detected speech segments.

Maintenance & Community

No specific details regarding maintainers, community channels, or project roadmap are present in the provided README.

Licensing & Compatibility

The specific open-source license for this project is not explicitly stated in the provided README text.

Limitations & Caveats

Installer builds must be performed natively on the target OS. Auto-switching is disabled by default when a specific microphone is set, requiring manual re-enabling. Devices with 'iphone' or 'continuity' keywords are ignored by default during auto-switching.