How to detect sentiment in a transcription?

Sentiment and Emotion Analysis

The Sentiment and Emotion Analysis feature enhances transcription results by detecting the sentiment (positive, negative, neutral, mixed, unknown) and emotion (fine-grained categories such as anger, joy, surprise, etc.) for each sentence in your audio.

When diarization is enabled, you can also analyze speaker-specific sentiments and emotions.

How to Enable

To activate sentiment and emotion analysis, set the sentiment_analysis parameter to true in your request.

Example Request


{
  "audio_url": "<your audio url>",
  "sentiment_analysis": true
}

Example Result

Each result entry includes:

text → spoken sentence
sentiment → overall polarity (positive/neutral/negative/mixed/unknown)
emotion → fine-grained emotional state
timestamps → start and end in seconds
channel / speaker → audio channel & speaker ID (if diarization enabled)


{
  "transcription": {...},
  "sentiment_analysis": {
    "success": true,
    "is_empty": false,
    "results": [
      {
        "text": "Jonathan, it says you are trained in technology.",
        "sentiment": "neutral",
        "emotion": "neutral",
        "start": 0.45158,
        "end": 2.364,
        "channel": 0,
        "speaker": 0
      },
      {
        "text": "That's very good.",
        "sentiment": "positive",
        "emotion": "positive_surprise",
        "start": 2.54438,
        "end": 3.54323,
        "channel": 0,
        "speaker": 0
      }
    ],
    "exec_time": 1.1271,
    "error": null
  }
}

Possible Values

Sentiment : positive, negative, neutral, mixed, unknown

Emotions:

Positive: adoration, amusement, awe, contentment, desire, ecstatic, elation, interest, positive_surprise, relief, sympathy, triumph
Negative: anger, confusion, contempt, disappointment, disgust, distress, embarrassment, fear, pain, sadness, negative_surprise
Neutral: realization, neutral