30 voiceovers of 45 seconds a month: 80 to 86 cents on Sume TTS

A 45-second voiceover costs about $0.026695 to $0.0285 on Sume TTS, so 30 of them in a month come to $0.80 to $0.86. Sume bills $0.0475 per 1,000 characters, and a spoken minute is roughly 750 to 800 characters, which is how Cartesia's pricing page converts credits to minutes.

The only wrinkle is rounding. The catalog quotes minimums in whole cents, so if each job rounds up to the next cent, a 562 to 600 character voiceover is 3 cents and 30 of them are $0.90. We show both because the exact-micro figure and the cent-rounded reserve are different numbers on different screens.

The calculation

Arithmetic

Characters = seconds / 60 x credits per minute. 45 / 60 = 0.75, and 0.75 x 750 = 562.5, rounded to 562; 0.75 x 800 = 600. Price = characters x $0.0000475, because Cartesia's list is 38 millionths of a dollar per credit, 1 credit is 1 character, and Sume multiplies by 1.25 (38 x 1.25 = 47.5 millionths). So 562 x $0.0000475 = $0.0267 and 600 x $0.0000475 = $0.0285.

For comparison, Cartesia's own Free plan, per its pricing page, includes 20K credits a month, which is 33 voiceovers of 600 characters. Sume has no free tier for TTS in the catalog, so the first 33 of your month are free on Cartesia directly and cost about $0.94 on Sume.

Steps

• Count the characters in the script, not the words. Billing is by transcript character.
• Submit on POST /v1/tts-1.0/generate for the managed voice path, or POST /v1/tts-router/generate with a Sonic id such as sonic-3.6 if you want to pin the engine.
• Keep a single request at or under 20,000 characters; longer scripts are split across requests.
• Multiply the characters by $0.0000475 before you submit. That is the budget line.

Scaling the same arithmetic

The cost is linear, so you can scale it without a new table. Ten times as many voiceovers (300 a month) at 600 characters is 300 x 600 x $0.0000475 = $8.55. A 90-second spot, which is 1,200 characters at the 800 credits-per-minute pace, is $0.057, exactly double the 45-second one. A 10-minute narration at 8,000 characters is $0.38.

The platform fee from Sume's pricing policy adds 5.5%, so the monthly 30-voiceover figure at 600 characters moves from $0.855 to $0.902. Even at the high end the voiceover line is the smallest item in almost any video budget: one 5-second Wan 3.0 clip at 720p costs more than 20 of these voiceovers.

That last comparison uses the same catalog: Wan 3.0 at 720p is $0.125 per second on Sume, so 5 seconds is $0.625, while one voiceover is $0.0285.

What Sume does not do

Sume does not bill by audio minute for TTS, does not give a free monthly allowance, and TTS Router carries Cartesia Sonic only for now. The 750 to 800 credits per minute is Cartesia's statement about its own audio and varies with language and speaking pace, so test one script before you extrapolate a month.