ProsodyAI Docs
TypeScript SDK

Analyze Audio

POST audio to api.prosodyai.app

Requires a dashboard psk_* key. See Integration.

const client = new ProsodyClient({
  apiKey: process.env.PROSODY_API_KEY!,
});

analyze()

await client.analyze('./recording.wav');
await client.analyze('https://example.com/audio.wav');
await client.analyze(audioBuffer);

Optional options: language, vertical, sessionId, includeFeatures.

analyzeBase64() / analyzePCM()

await client.analyzeBase64(base64Audio);
await client.analyzePCM(float32Samples, { sampleRate: 16000 });

Request (multipart)

POST https://api.prosodyai.app/v1/analyze/audio
X-API-Key: psk_...
Content-Type: multipart/form-data

file=<wav>

Response

result.valence;
result.arousal;
result.dominance;
result.signals;
result.turns; // diarized turns with turn.prosody (interview product)
result.kpi_predictions; // when KPIs are configured for your org

On this page