Text to speech API price per 1,000 characters in October 2026
From $0.011 (Eleven v4 Turbo, promo) to $0.08 (Eleven v3) per 1K characters, with Deepgram, OpenAI and Sume's Sonic at $0.0475 in one dated table.

On the price pages read on 2026-10-05, character-priced text to speech runs from $0.011 per 1,000 characters (Eleven v4 Turbo, in a promo through Oct 12) up to $0.08 (Eleven v3). The middle of the field sits at $0.015 to $0.045. Sume's Sonic route is $0.0475 per 1K: Cartesia's list of $38 per 1M credits, with one credit per character, times 1.25 (47.5 micro-dollars per character), before the 5.5% platform fee.
Price is not quality. A cheap model that needs two takes is not cheap, and models differ in languages, voices and controls. Use the table to set a budget, then audition.
Price per 1,000 characters, sorted
OpenAI and Deepgram list per million or per 1K characters; the figures here are converted to per 1K. For example $15 per 1M characters is $0.015 per 1K.
| Model | Per 1K characters | Note |
|---|---|---|
| Eleven v4 Turbo | $0.011 | Promo, through Oct 12 |
| OpenAI tts-1 | $0.015 | $15 per 1M characters |
| Deepgram Aura-1 | $0.015 | |
| Eleven v4 | $0.022 | Promo, 72% off until Oct 12 |
| Deepgram Aura-2 | $0.030 | |
| OpenAI tts-1-hd | $0.030 | $30 per 1M characters |
| Cartesia Sonic at list | $0.038 | $38 per 1M credits, 1 credit = 1 character |
| Eleven Flash/Turbo | $0.04 | |
| Deepgram Flux TTS | $0.045 | |
| Sonic-3.6 via Sume | $0.0475 | List x1.25 = 47.5 micros per character, before fee and cent rounding |
| Eleven v3 | $0.08 |
What the Sume number includes
Sume bills provider list x1.25 across SKUs that have a provider list, and a 5.5% platform fee on customer spend (policy live since 2026-09-22). For TTS the character rate is $0.038 x 1.25 = $0.0475 per 1K, with each job rounded up to the cent. A 20,000-character job, the maximum, is exactly $0.95 before fee. A 1,000-character job is $0.0475, billed as $0.05.
Sume only serves Cartesia Sonic here. It does not resell Eleven, OpenAI or Deepgram voices, so rows other than the last Sonic ones are references, not Sume prices.
Reading the table honestly
- Promos end. Eleven's v4 prices are marked until Oct 12 and the page text does not give the post-promo price.
- Gemini TTS is priced per token, not per character, so it is not in this table; see the separate Gemini pricing post.
- Characters are not seconds. Cartesia states 750-800 characters per audio minute; other vendors do not publish an equivalent in the pages read.
- Per-job rounding matters for short reads: at $0.0475 per 1K, a 200-character line is $0.0095 but bills at one cent.
Sources
Related posts
More in Pricing
- Envato Elements Core vs AI music at $0.125: the 132-track break-even
Envato Elements Core is $16.50 a month on annual billing. At $0.125 per Sume track, you break even near 132 tracks a month. See the math and what it ignores.
- Estimate a Sume video request first: rate x seconds x 1.25
Per-second models bill list price times 1.25 times seconds. A small Python estimator, three 10-second examples and how reserve and usage.cost relate.
- Every MiniMax price on Sume: H3, H3 Max, lip sync, Recast
One table of the per-second prices Sume bills for MiniMax H3, H3 Max, H3 Max lip sync and H3 Max Recast at each tier, list x 1.25, with 10-second totals.
- Faceless channel, 16 videos a month: the Sume wallet budget
Four 5-minute faceless videos a week cost about $5.29 each on Sume, so $85 a month. How to size top-ups and a retry reserve against 365-day credit lots.
Written by Sume