# geko

> Text-to-speech API for Kazakh — and beyond. A thin, familiar SDK and a simple HTTP API.

## Docs

- [List models](https://docs.geko.sh/api-reference/catalog/list-models.md): List the models on the platform and their voices. **Open** — no API key required.
- [List voices](https://docs.geko.sh/api-reference/catalog/list-voices.md): The voice catalog for a model. **Open** — no API key required.
- [Service status](https://docs.geko.sh/api-reference/catalog/service-status.md): Liveness probe. **Open** — no API key required.
- [OpenAI-compatible speech](https://docs.geko.sh/api-reference/speech/openai-compatible-speech.md): OpenAI-compatible alias — point the OpenAI SDK at geko by swapping the base URL and model. Maps OpenAI's request shape onto the Tokay engine. See the [OpenAI compatibility guide](/guides/openai).
- [Stream speech sentence-by-sentence](https://docs.geko.sh/api-reference/speech/stream-speech-sentence-by-sentence.md): Same request body as `POST /v1/tts`, but the response is delivered as each sentence finishes synthesizing — lower time-to-first-audio on long text. Auth and metering are identical to `/v1/tts`.
- [Synthesize speech](https://docs.geko.sh/api-reference/speech/synthesize-speech.md): Synthesize text into speech. Returns raw **WAV** bytes (`audio/wav`, 24 kHz, 16-bit PCM, mono). The `X-Tokay-Chars` response header reports the billed character count.
- [Overview](https://docs.geko.sh/api/reference.md): Base URL, auth, conventions, and errors for the geko HTTP API.
- [Authentication](https://docs.geko.sh/authentication.md): Creating and using API keys, and keeping them safe.
- [Billing & credits](https://docs.geko.sh/billing.md): How Geko meters usage, what it costs, and how credits work.
- [Changelog](https://docs.geko.sh/changelog.md): Notable changes to the Geko API and SDK.
- [FAQ & troubleshooting](https://docs.geko.sh/faq.md): Common questions and fixes.
- [Use it in your app](https://docs.geko.sh/guides/frameworks.md): Server-side integration for Next.js and Express — the safe way.
- [Latency & quality (NFE)](https://docs.geko.sh/guides/latency.md): Tune the nfe dial and understand cold starts.
- [Text normalization](https://docs.geko.sh/guides/normalization.md): How Geko turns numbers, currency, and dates into spoken Kazakh.
- [OpenAI compatibility](https://docs.geko.sh/guides/openai.md): Point the OpenAI SDK at Geko by changing two lines.
- [Playing & saving audio](https://docs.geko.sh/guides/playing-audio.md): Turn the WAV bytes into files, playback, and other formats.
- [Streaming audio](https://docs.geko.sh/guides/streaming.md): Start playing sentence one while the rest still synthesizes.
- [Voice agents](https://docs.geko.sh/guides/voice-agents.md): Use Geko as the low-latency voice for a phone or assistant loop.
- [text-to-speech that just works](https://docs.geko.sh/index.md): send text, get natural speech in Kazakh and beyond. a thin SDK, a CLI, and a simple HTTP API.
- [Limits & constraints](https://docs.geko.sh/limits.md): Request limits, output format, rate limits, and cold starts.
- [Quickstart](https://docs.geko.sh/quickstart.md): From zero to your first hello.wav in three steps.
- [Roadmap](https://docs.geko.sh/roadmap.md): What's planned but not yet available.
- [CLI](https://docs.geko.sh/sdk/cli.md): Synthesize speech and browse the catalog from your terminal — no code.
- [Errors & retries](https://docs.geko.sh/sdk/errors.md): Typed errors, automatic retries, timeouts, and cancellation.
- [TypeScript / JavaScript SDK](https://docs.geko.sh/sdk/typescript.md): @gekoai/sdk — a thin, typed, zero-dependency client for the Geko API.
- [Voices](https://docs.geko.sh/voices.md): The six Kazakh voices of tokay-kk-v1 — and how to audition them.

## OpenAPI Specs

- [openapi](https://docs.geko.sh/openapi.json)