TTS API price per 1,000 characters: Sume, MAI, Groq, Lemonfox, Unreal
A per-1,000-character table of list prices for Sume, MAI-Voice-2.1 and Flash, Groq Orpheus, Lemonfox and Unreal Speech, with cost for a 750-character short.
Text-to-speech APIs quote prices in different units: per million characters, per credit, per monthly plan. Converting everything to dollars per 1,000 characters makes them comparable. The table below does that from each vendor's own page (read 2026-10-06), and adds the cost of one 750-character narration, roughly a minute of speech.
Sume's rate is the highest in the table, and the sections below say what it covers and what it does not.
List prices side by side
| Service | Per million characters | Per 1,000 characters | Source |
|---|---|---|---|
| Sume TTS 1.0 | $47.50 | $0.0475 | Sume docs and catalog |
| MAI-Voice-2.1 | $22.00 | $0.0220 | Digital Applied tracker |
| MAI-Voice-2.1-Flash | $15.00 | $0.0150 | Digital Applied tracker |
| Groq Orpheus V1 English | $22.00 | $0.0220 | Groq model page |
| Lemonfox (derived) | $2.50 | $0.0025 | Lemonfox page |
| Unreal Speech Plus (derived) | $11.88 | $0.0119 | Unreal pricing page |
One 750-character narration
| Service | Cost of 750 characters |
|---|---|
| Sume TTS 1.0 | $0.04 (exact $0.03562) |
| MAI-Voice-2.1 | $0.01650 |
| MAI-Voice-2.1-Flash | $0.01125 |
| Groq Orpheus V1 English | $0.01650 (4 requests at 200 characters) |
| Lemonfox (derived) | $0.00187 |
Reading the numbers
Microsoft's two voice models were announced on 2026-10-01 at $22 and $15 per million characters; the tracker also lists 23 languages and 26 locales (Digital Applied, PYMNTS). All of these are list prices before tax, minimums, taxes or volume deals. The vendors' own billing granularity may round up, which this table does not model.
On a single 750-character short, even the largest gap in the table is under five cents. The gap matters at volume, so price the month, not the clip: 3,000 shorts at 750 characters is 2,250,000 characters, or $120.00 on Sume against $49.50 at the MAI rate.
Sources
- Digital Applied: AI Model Releases, October 2026 tracker (read 2026-10-06)
- PYMNTS: Microsoft targets contact center market with real-time AI voice models (read 2026-10-06)
- Groq Docs: Orpheus V1 English (read 2026-10-06)
- Lemonfox.ai: Text-to-speech API (read 2026-10-06)
- Unreal Speech: Pricing (read 2026-10-06)
- Sume API reference
Related posts
More in Pricing
- Transcribe 100,000 two-second voice command recordings: the bill
A two-second clip costs about $0.000333 on Sume STT, so 100,000 cost $33.33. Without duration_seconds each job in flight holds a full $0.01 instead.
- Trim first, then caption: the 60-second note on the caption price
Sume quotes the $0.20 video-captions price for videos up to 60 seconds. Cut the episode with video trim at $0.02 first, then caption only the part you keep.
- 52 weekly newsletter header images: Ideogram 4.5 cost by tier on Sume
A year of weekly newsletter headers on Ideogram 4.5: $1.95 on low, $3.90 on medium, $14.30 on high, plus a three-drafts-then-one-final plan at $20.15.
- What one captioned Shorts episode costs on Sume, itemized
Add the documented Sume rates for one Shorts episode: render, trim, captions, transcript. A Python estimator refuses captions past the 60-second quote.
Written by Sume