Openai Whisper

Openai Whisper

Playground

Whisper Large v3 speech-to-text: instant, accurate multilingual transcripts with automatic language detection and punctuation. Upload audio to get transcripts. Ready-to-use REST API, no coldstarts, affordable pricing.

Features

OpenAI Whisper (Large-v3) is a production-ready speech recognition model that transcribes or translates audio into clean, readable text. With support for dozens of languages, optional word-level timestamps, and flexible task modes — it’s ideal for subtitling, transcription, and multilingual workflows.

Why It Stands Out

• Transcribe or translate: Choose between same-language transcription or translation to English.
• Multilingual support: Transcribe audio in dozens of languages with automatic language detection.
• Word-level timestamps: Generate precise timing data for subtitle alignment and editing workflows.
• Prompt-guided output: Steer formatting, terminology, or punctuation with custom prompts.
• Prompt Enhancer: Built-in AI-powered prompt optimization for better transcription guidance.
• Flexible input: Supports direct audio upload or public URL.

Parameters

How to Use

1. Upload your audio — drag and drop a file or paste a public URL.
2. Select language — choose a specific language or use “auto” for detection.
3. Choose task — select “transcribe” for same-language output or “translate” for English.
4. Enable timestamps (optional) — turn on for word-level timing data.
5. Add a prompt (optional) — guide formatting, punctuation, or specific terminology.
6. Click Run and wait for transcription to complete.
7. Copy or download the transcribed text.

Best Use Cases

• Subtitle Generation — Create accurate, timed transcripts for video subtitling.
• Meeting Transcription — Convert recorded meetings, interviews, and calls into searchable text.
• Translation Workflows — Translate foreign-language audio directly to English text.
• Content Repurposing — Turn podcasts, webinars, and lectures into written content.
• Accessibility — Generate transcripts to make audio content accessible.

Pricing

Total cost = duration of audio (in seconds) × price per second

Examples

• 60s audio (standard) → 60 × $0.001 = $0.06
• 60s audio (with timestamps) → 60 × $0.002 = $0.12
• 10 min (600s) audio (standard) → 600 × $0.001 = $0.60
• 10 min (600s) audio (with timestamps) → 600 × $0.002 = $1.20

Pro Tips for Best Quality

• Use clear audio with minimal background noise for optimal accuracy.
• Specify the language manually if auto-detection is inconsistent.
• Enable timestamps only when needed for subtitles or alignment — it doubles the cost.
• Add a prompt to guide transcription — include names, jargon, or formatting preferences.
• Use “translate” task for non-English audio when you need English output.

Notes

• Ensure uploaded audio URLs are publicly accessible.
• Timestamps are best for subtitles and editing, but may take longer to process.
• Processing time varies based on audio duration and current queue load.

Duration limit

The maximum supported audio duration is 10 minutes. Longer audio is automatically trimmed to 10 minutes before processing.

Authentication

For authentication details, please refer to the Authentication Guide.

API Endpoints

Submit Task & Query Result

set -euo pipefail

export WAVESPEED_API_KEY="your-api-key"

REQUEST_BODY=$(cat <<'JSON'
{
  "audio": "https://interactive-examples.mdn.mozilla.net/media/cc0-audio/t-rex-roar.mp3",
  "language": "auto",
  "task": "transcribe",
  "enable_timestamps": false
}
JSON
)

# 1. Submit the prediction.
SUBMIT_RESPONSE=$(curl --silent --show-error --fail-with-body \
  -X POST "https://api.wavespeed.ai/api/v3/wavespeed-ai/openai-whisper" \
  -H "Authorization: Bearer ${WAVESPEED_API_KEY}" \
  -H "Content-Type: application/json" \
  -d "${REQUEST_BODY}")

TASK=$(printf '%s' "${SUBMIT_RESPONSE}" | jq 'if type == "object" and has("data") then .data else . end')
PREDICTION_ID=$(printf '%s' "${TASK}" | jq -r '.id // empty')
if [ -z "${PREDICTION_ID}" ]; then
  printf 'Submission response did not contain a prediction id
' >&2
  exit 1
fi
RESULT_URL=$(printf '%s' "${TASK}" | jq -r '.urls.get // empty')
if [ -z "${RESULT_URL}" ]; then RESULT_URL="https://api.wavespeed.ai/api/v3/predictions/${PREDICTION_ID}/result"; fi

# 2. Poll until the prediction finishes.
while true; do
  RESPONSE=$(curl --silent --show-error --fail-with-body \
    "${RESULT_URL}" \
    -H "Authorization: Bearer ${WAVESPEED_API_KEY}")
  RESULT=$(printf '%s' "${RESPONSE}" | jq 'if type == "object" and has("data") then .data else . end')
  STATUS=$(printf '%s' "${RESULT}" | jq -r '.status // empty')

  case "${STATUS}" in
    completed) printf '%s\n' "${RESULT}" | jq '.outputs'; break ;;
    failed|cancelled|timeout) printf '%s\n' "${RESULT}" | jq . >&2; exit 1 ;;
    created|processing) sleep 2 ;;
    *) printf 'Unexpected status: %s
' "${STATUS}" >&2; exit 1 ;;
  esac
done

Parameters

Task Submission Parameters

Request Parameters

Response Parameters

Result Request Parameters

Result Response Parameters