Kokori

Product Introduction

1. Definition: Kokori is a macOS-native text-to-speech (TTS) application and local API server that converts text into high-quality audio entirely offline. It falls under the technical category of desktop TTS software with embedded REST API capabilities.
2. Core Value Proposition: Kokori eliminates dependency on cloud-based TTS services by providing offline, privacy-focused speech synthesis with studio-grade voices, reducing costs and latency for developers and creators.

Main Features

1. Local REST API Server:

   • How it works: Runs a lightweight HTTP server (port 5002) on your Mac. Send JSON payloads via POST requests (text, voice, speed) to generate audio without internet.
   • Technology: Built on Kokoro TTS engine, leveraging neural networks for natural prosody. Processes requests locally, avoiding cloud dependencies.

2. Multi-Voice Library:

   • How it works: Offers 50+ preloaded voices across 8 languages (e.g., American/British English, Japanese, Mandarin). Voices are quality-ranked (A-F) for clarity.
   • Technology: Utilizes optimized acoustic models for each voice, supporting gender/language filters (e.g., en_us_heart for high-quality American female).

3. Audio History & Logging:

   • How it works: Automatically archives generated audio files locally. Detailed logs track API requests, errors, and performance metrics (e.g., latency).
   • Technology: File-based storage with timestamped entries, enabling debugging without third-party tools.

Problems Solved

1. Pain Point: High costs and privacy risks of cloud TTS APIs (e.g., Google Cloud, AWS Polly). Kokori enables $0 operational expenses and zero data leakage.
2. Target Audience:
   • Developers: Test voice-enabled apps offline, avoiding per-request fees.
   • Content Creators: Generate unlimited voiceovers for videos/podcasts without subscriptions.
   • Privacy-Conscious Users: Process sensitive documents offline (e.g., legal/medical text).
3. Use Cases:
   • Prototyping voice assistants without API keys.
   • Creating multilingual audiobooks offline.
   • Accessibility tools for offline text consumption.

Unique Advantages

1. Differentiation:
   • Vs. Cloud TTS: No throttling, 100% offline, no recurring fees.
   • Vs. Built-in macOS TTS: Higher-quality voices, developer API, and speed/pitch control.
2. Key Innovation:
   • Integrated Desktop-App/API Hybrid: Seamlessly switch between GUI (menubar) and programmatic use.
   • Voice Quality Hierarchy: Curated voice library with transparency about performance (e.g., "A-grade" vs. "D-grade" voices).

Frequently Asked Questions (FAQ)

1. Does Kokori work without an internet connection?
   Yes, Kokori’s TTS engine and API server run entirely offline—no data leaves your device.

2. Can I use Kokori voices for commercial projects?
   Absolutely. The license permits commercial usage, including video monetization and app integration.

3. How resource-intensive is the local API server?
   Optimized for efficiency: uses <500MB RAM during operation and supports concurrent requests on modern Macs.

4. What languages and accents does Kokori support?
   Includes 8 languages (English, Japanese, Spanish, etc.) with regional variants like British (bf_alice) and American (af_heart) voices.

5. Is there a Windows or Linux version?
   Currently, Kokori is exclusive to macOS, leveraging native Apple Silicon/Intel optimizations.