richer-richard/cochlea

Overview

A headless, deterministic audio engine for AI agents. It composes scores as data, renders byte-reproducible PCM offline, and enables analysis through features, spectrograms, and assertions without requiring an audio device or human ear.

Best for

Best for
Developers building AI agents that need deterministic, testable audio generation

Use cases

• Generate deterministic audio samples for testing AI audio models
• Create reproducible soundscapes for simulation environments
• Validate audio output programmatically with assertions

Notes

A headless, deterministic audio engine for AI agents. It composes scores as data, renders byte-reproducible PCM offline, and enables analysis through features, spectrograms, and assertions without requiring an audio device or human ear.

0 stars on GitHub. Last updated 2026-07-11. Licensed Apache-2.0.

Use cases

• Generate deterministic audio samples for testing AI audio models
• Create reproducible soundscapes for simulation environments
• Validate audio output programmatically with assertions

Pros

• Deterministic output ensures reproducibility across runs
• No dependency on audio hardware or external tools like ffmpeg
• Written in Rust for performance and safety

Cons

• Limited to offline rendering, no real-time audio processing
• Small community with zero stars, may lack support or documentation
• Niche use case focused on AI agents, not general audio applications

Indexed from awesome-mcp-servers-punkpeye and enriched against its public facts.