ElevenLabs dubbing $0.33, $0.50 and $2.20 per minute vs Sume steps

ElevenLabs' API pricing page lists dubbing v1 at $0.33 per minute with a watermark and $0.50 per minute without, and dubbing v2 at $2.20 per minute. Sume does not sell dubbing as one SKU in its docs; you build it from steps whose public prices are: speech-to-text at $0.01 per audio minute, audio detach at $0.01 per job, timeline audio at $0.01 flat and captions at $0.20 per job for videos up to 60 seconds. Translation and the TTS step are not priced in the docs.

Price lines

ElevenLabs' column is from its pricing page; Sume's rows are from its docs. The TTS and translation lines are intentionally blank because no public Sume rate exists for them in the docs.

Where Sume's steps start

Detach audio from the video, then transcribe with POST /v1/stt-1.0/transcribe with sentence segmentation to keep timing for each line.

import os, requests

r = requests.post(
    "https://api.sume.com/v1/stt-1.0/transcribe",
    headers={"Authorization": f"Bearer {os.environ['SUME_API_KEY']}",
             "Idempotency-Key": "stt-demo-001"},
    json={"audio_url": "https://example.com/call.mp3",
          "language_code": "en",
          "duration_seconds": 300},
)
print(r.status_code, r.json())

Which to choose

If you want one call per video, ElevenLabs' bundled product is the shorter path and its prices are listed above. If you need control per step, such as editing the translation before speaking it, the Sume steps give you that. Price a one-minute sample both ways before committing.

Worked example

Here is a 10-minute video priced from the published lines.

• ElevenLabs v1 watermarked: $3.30; v1 clean: $5.00; v2: $22.00.
• Sume's priced steps for the same 10 minutes: STT $0.10, audio detach $0.01, timeline audio $0.01, which is $0.12 before translation and speech.
• Translation and TTS prices are not in Sume's docs, so a full total is not given here.

Checklist before you commit

A fair comparison needs a real Sume test job for the speech step, plus your translation choice.

• Test lip-sync needs separately.
• Check watermark rules on v1.
• Keep the original audio for fixes.