Audio speech (TTS)

Generate audio from text. OpenAI-compatible.

POST https://api.ecohash.com/v1/audio/speech

Headers

Header	Value
Authorization	Bearer eco_YOUR_KEY
Content-Type	application/json

Request body

Field	Type	Required	Notes
model	string	yes	TTS model ID (e.g. kokoro-82m)
voice	string	yes	Voice identifier supported by the model
input	string	yes	Text to synthesize, up to 4000 characters
response_format	string	no	"mp3" (default), "wav", or "opus"
speed	number	no	0.5–2.0 playback speed multiplier
stream	boolean	no	When true, audio bytes are streamed as they are generated instead of buffered until the end — lower time-to-first-audio for long inputs

Example request

curl https://api.ecohash.com/v1/audio/speech \
  -H "Authorization: Bearer eco_YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "kokoro-82m",
    "voice": "af_bella",
    "input": "Hello from EcoLink."
  }' \
  --output hello.mp3

Response

The response body is the audio bytes directly, with Content-Type matching the requested format (audio/mpeg, audio/wav, or audio/ogg). Pipe directly to a file.

Streaming

Set "stream": true to receive the audio as a stream instead of a single buffered response. The bytes begin arriving as soon as the model starts producing them, which lowers time-to-first-audio for longer text. Write the chunks to a file or feed them to an audio player as they arrive:

curl https://api.ecohash.com/v1/audio/speech \
  -H "Authorization: Bearer eco_YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "kokoro-82m",
    "voice": "af_bella",
    "input": "This response is streamed as it is generated.",
    "stream": true
  }' \
  --output stream.mp3

Errors

See Errors. Common: 401, 402, 404, 429.