Speechmatics Melia 1 $0.12 per hour vs Sume STT $0.60 per hour

Speechmatics' pricing page lists its Melia 1 model at $0.12 per hour across 56 languages, with Enhanced at $0.38 per hour. Sume STT 1.0 is $0.01 per audio minute, or $0.60 per hour. Speechmatics is cheaper than Sume on every tier listed here.

Published Speechmatics rates

The page also lists a translation add-on and free credits.

Using Sume's STT

Send audio_url, optional language_code and a duration_seconds hint (1 to 600). The result returns text plus words[] timings.

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())

Honest summary

On price per hour, Speechmatics undercuts Sume. Sume's draw is that the transcript lives in the same job system as captions, dubbing and timeline audio. Run a sample, compare accuracy on your audio, and weigh the cost of a second vendor integration.

Worked example

The page lists several models, so the comparison depends on which you would actually pick.

• 100 hours: Melia 1 $12.00, Oak 1 $15.00, Standard $23.00, Enhanced $38.00, Sume STT $60.00.
• Speechmatics' $100 free credit alone covers 833 hours of Melia 1 at $0.12 per hour, before any paid use.
• Speechmatics lists translation at $0.49 per hour as an add-on.

Checklist before you commit

If you need translated transcripts, price that add-on in: Sume's docs describe translation as a separate step you add; STT returns text and word timings.

• Pick the Speechmatics model your languages need.
• Confirm the Sume job fits your file lengths (600-second hint cap).
• Note that both prices can change after 2026-10-02.