Docs

Quickstart

This guide starts with the OpenAI-compatible API because it is the primary integration surface for Oolel Speech V1.

1. Create An API Key

Open the Soynade dashboard, choose your workspace, and create an API key for the project you want to use.

Keep the key secret. Use it from a server, CLI, backend job, or trusted testing environment. Do not ship it inside browser code or mobile apps.

2. Export The Key

Code
export SOYNADE_API_KEY="zpka_..."

3. Use The OpenAI-Compatible API

Install the OpenAI Python SDK if you use Python:

Code
pip install openai

Then use your Soynade API key with Soynade's OpenAI-compatible base URL.

For speech recognition, keep temperature at 0.

4. Go Deeper

Use the pages below for the full examples and operational details:

Page	Use when
Models	You want model capabilities and examples for Oolel Speech V1 and Oolel Voices.
OpenAI SDK clients	You want text translation, speech recognition, speech translation, and streaming through OpenAI-compatible clients.
Task-Specific APIs	You want dedicated endpoints for text translation, speech translation, speech recognition and text to speech.
Async Jobs	You need polling, download URLs, webhooks, or longer-running requests.
Voice Rights And Safety	You use reference audio for text to speech.
Pricing	You want metering and rate information.
API Reference	You want schemas, request fields, response shapes, and the interactive playground.

Last modified on September 18, 2026

Pricing

3. Use The OpenAI-Compatible API

Install the OpenAI Python SDK if you use Python:

Code

pip install openai

Then use your Soynade API key with Soynade's OpenAI-compatible base URL.

For speech recognition, keep temperature at 0.

AUDIO_BASE64="$(base64 -i ./audio.wav | tr -d '\n')"

cat > /tmp/speech-recognition-request.json <<EOF
{
  "model": "oolel-speech-v1",
  "modalities": ["text"],
  "temperature": 0,
  "max_tokens": 256,
  "messages": [
    {
      "role": "user",
      "content": [
        {
          "type": "input_audio",
          "input_audio": {
            "data": "$AUDIO_BASE64",
            "format": "wav"
          }
        },
        {
          "type": "text",
          "text": "Transcribe this audio. Reply with the transcription only."
        }
      ]
    }
  ]
}
EOF

curl -X POST "https://api.soynade.ai/openai/v1/chat/completions" \
  -H "Authorization: Bearer $SOYNADE_API_KEY" \
  -H "Content-Type: application/json" \
  -d @/tmp/speech-recognition-request.json