![\"Kesha](\"https://github.com/drakulavich/kesha-voice-kit/raw/main/docs/assets/logo.png\"){kind=logo}
\n\n \n
\n ![\"Tests\"](\"https://img.shields.io/endpoint?url=https%3A%2F%2Fflakiness.io%2Fapi%2Fbadge%3Finput%3D%257B%2522badgeToken%2522%253A%2522badge-2IKMRRqUxh9P3w8Ym3Szf0%2522%257D\")
![\"npm](\"https://img.shields.io/npm/v/@drakulavich/kesha-voice-kit\")
![\"License:](\"https://img.shields.io/badge/License-MIT-blue.svg\"){kind=icon}
\n ![\"Bun\"](\"https://img.shields.io/badge/runtime-Bun-f9f1e1?logo=bun\"){kind=icon}
Give your local tools and LLM agents a voice.
Fast speech-to-text, text-to-speech, voice-activity detection, and language detection in one local-first CLI — CoreML on Apple Silicon, ONNX on Linux and Windows.
\n ![\"kesha](\"https://github.com/drakulavich/kesha-voice-kit/raw/main/docs/assets/demo.webp\")
\n
` (or the voice with `--voice `) — otherwise the engine default speaks.\n\n```bash\nkesha install --tts # English voices; sizes differ per platform — preview: kesha install --plan\nkesha install --tts en ru # + Russian (+~890 MB, Vosk)\nkesha say \"Hello, world\" > hello.wav\nkesha say \"Привет, мир\" > privet.wav # auto-routes by language (macOS)\nkesha say --lang ru \"Привет, мир\" > privet.wav # explicit — the Linux/Windows path\nkesha say --voice ru-vosk-m02 \"Голос в текст.\" > ru.wav\n```\n\n**Output formats** (`--format`, or inferred from the `--out` extension):\n\n```bash\nkesha say \"Hello\" --out hi.wav # WAV (default, uncompressed)\nkesha say \"Hello\" --format ogg-opus --out hi.ogg # OGG/Opus — messenger voice notes\nkesha say \"Hello\" --format flac --out hi.flac # FLAC — lossless, plays in every browser incl. Safari/iOS\n```\n\n`kesha say --list-voices` lists what's installed. Voices, the full catalogue, macOS system voices, SSML, speaking rate (`--rate`, ``), Russian word stress, and Russian/English abbreviation handling are all in **[docs/tts.md](docs/tts.md)**.\n\n## Languages\n\n**Speech-to-text** spans 25 languages and **text-to-speech** 9 — full tables with codes, flags, and per-platform availability in **[docs/languages.md](docs/languages.md)**. Audio language detection identifies [107 languages](https://huggingface.co/speechbrain/lang-id-voxlingua107-ecapa).\n\n## Performance\n\n> **Up to ~19x faster than Whisper** on Apple Silicon (M2), **~2.5x faster** on CPU\n\nCompared against Whisper `large-v3-turbo`, all engines auto-detecting language:\n\n\n\nFull per-file breakdown (Russian + English): [BENCHMARK.md](BENCHMARK.md). The CPU figure is the ONNX engine on an M2's CPU cores; no x86 numbers are published yet.\n\n## Other install methods\n\nAll of these install the Bun CLI wrapper; engine + models still download explicitly via `kesha install`. (Nix is the exception — it currently builds only the engine from source; see below.)\n\n- **Homebrew** — `brew install drakulavich/tap/kesha-voice-kit` · [docs/distribution.md](docs/distribution.md#homebrew)\n- **Linux packages** (`.deb`/`.rpm`, x64) — published on stable releases, see [docs/distribution.md](docs/distribution.md#linux-packages)\n- **Docker** (GHCR image) — [docs/docker.md](docs/docker.md)\n- **Nix** (`aarch64-darwin` / `x86_64-linux`) — builds the engine from source (`nix build github:drakulavich/kesha-voice-kit#kesha-engine`). The full `kesha` CLI via `nix run` / `nix profile install` is **not yet available** — it needs a maintainer with Nix to populate a build hash ([#946](https://github.com/drakulavich/kesha-voice-kit/issues/946)). · [docs/distribution.md](docs/distribution.md#nix)\n- **Shell completions + manpage** — `kesha completions bash|zsh|fish` and `kesha manpage` print the packaged files to install wherever your shell expects them.\n\n## Integrations\n\n- **MCP server** — `kesha mcp` exposes transcribe/synthesize/list tools to any MCP client (Claude, Cursor, Codex, Gemini). Setup: [docs/mcp.md](docs/mcp.md).\n- **OpenClaw** — give your LLM agent ears. Install & config: [docs/openclaw.md](docs/openclaw.md).\n- **Hermes Agent** — local STT/TTS through Hermes command providers. Setup: [docs/hermes.md](docs/hermes.md).\n- **Raycast** (macOS) — offline microphone dictation from the launcher: *Dictate to Clipboard* records with a live signal meter, auto-stops on silence, transcribes locally, and copies the text. [Install from the Raycast Store](https://www.raycast.com/drakulavich/kesha-voice-kit) · source: [`raycast/`](raycast/).\n- **Programmatic API** — `@drakulavich/kesha-voice-kit/core` for use inside a Bun program. See [docs/api.md](docs/api.md).\n\n## More\n\n- [Architecture](docs/architecture.md) — runtime data flow, the models that ship, the CLI ↔ Rust engine boundary, model pinning, and where tests live.\n- [Use cases](docs/use-cases.md) — copy-paste recipes (transcribe a meeting, speak from OpenClaw, run offline, move the cache).\n- [Product positioning](docs/product-positioning.md) — supported workflows, non-goals, maturity labels, platform matrix.\n- [Changelog](CHANGELOG.md) — every release, with the behaviour changes spelled out.\n- **Diagnostics:** `kesha doctor`, `kesha support-bundle` (redacted `.tar.gz` for issues), and `kesha logs` produce local, content-free diagnostics — see [docs/diagnostic-logs.md](docs/diagnostic-logs.md). One variable, `KESHA_HOME`, moves everything Kesha writes (models, logs, stats) into a directory of your choice — see [where Kesha keeps its files](docs/diagnostic-logs.md#where-kesha-keeps-its-files). Every failure prints a stable `error [CODE]: …` line and a documented [process exit code](docs/errors.md#process-exit-codes).\n- **Scripting & CI:** `--json` (or `--toon`) for machine-readable output, `--include-errors` (with either) to get per-file failures on stdout alongside the results, `--quiet`/`-q` to silence progress, and `--no-color` (or `NO_COLOR=1`) for plain logs. Colors switch off automatically when `CI=true`.\n- **Privacy / Local Stats:** Stats are **off by default** and fully local. Opt in with `kesha stats enable` to record content-free operational metrics in a local SQLite database — never networked, never storing audio, transcripts, text, or paths. Full commands & lifecycle: [docs/local-stats.md](docs/local-stats.md).\n\n## Contributing\n\nSee [CONTRIBUTING.md](CONTRIBUTING.md), the [Roadmap](ROADMAP.md) (Now / Next / Later), and the [Decision log](docs/decision-log.md) (why platform/model choices were made — and reversed). Dev setup: `just dev-setup` (Bun, Rust, nextest, platform libs).\n\n## License\n\nMade with 💛🩵 and 🥤 energy under MIT License\n","readmeFilename":"README.md"}