@absolutejs/voice-neets

Neets text-to-speech adapter for @absolutejs/voice.

Wraps Neets's /v1/tts HTTP endpoint for low-cost TTS using their ar-diff-50k / style-tts-2 / vits models. Streams audio body straight through to voice's runtime.

Install

bun add @absolutejs/voice-neets

@absolutejs/voice is a runtime dependency.

Use

import { voice } from "@absolutejs/voice";
import { neets } from "@absolutejs/voice-neets";

const app = voice({
  tts: neets({
    apiKey: process.env.NEETS_API_KEY!,
    voiceId: "us-male-2",
    // optional:
    model: "ar-diff-50k",      // default; or 'style-tts-2' / 'vits'
    format: "pcm",             // default; only "pcm" is supported for streaming
    sampleRate: 22_050,        // default
    language: "en-us",
    temperature: 0.7,
  }),
});

Options

| Option | Required | Default | Notes | | --- | --- | --- | --- | | apiKey | yes | — | Neets API key, sent as X-API-Key. | | voiceId | yes | — | Neets voice id. | | model | no | ar-diff-50k | ar-diff-50k, style-tts-2, vits, or a future Neets model id. | | format | no | pcm | Must be pcm. mp3 / wav are rejected because they aren't streamable frame-by-frame. | | sampleRate | no | 22050 | Forwarded as sample_rate. | | language | no | — | Forwarded as language (e.g. en-us, es-es). | | temperature | no | — | Forwarded inside params. | | baseUrl | no | https://api.neets.ai | Override for staging. | | fetch | no | globalThis.fetch | Inject for tests; opportunistic HTTP/2 multiplexing on outbound HTTPS. |

Notes

• Whitespace-only send() is a no-op.
• session.close(reason) aborts in-flight requests and refuses further sends.