Cartesia Startup and Scale vs Sume TTS: break-even in characters

Sume TTS bills $0.0475 per 1,000 characters with no monthly fee. Cartesia's own Startup plan works out to $0.0392 per 1,000 credits and Scale to $0.037375, so a subscription is cheaper per character only if you use nearly all of it: the break-even is about 1,031,579 characters a month for Startup and 6,294,737 for Scale, before Sume's platform fee.

Cartesia's pricing page, read today, lists Free at $0 for 20K credits a month, Pro at $5 for 100K, Startup at $49 for 1.25M, Scale at $299 for 8M, and Enterprise on request. It states that one credit equals one character for text-to-speech and that a minute of audio takes 750 to 800 credits. Sume's TTS 1.0 and TTS Router use the same one-credit-per-character rule on Sonic.

Per-character comparison

The arithmetic

Sume's list per credit in the repo is 38 millionths of a dollar, so 1,000 characters is $0.038 at provider list and $0.038 x 1.25 = $0.0475 billed. Cartesia Startup: $49 / 1,250,000 = $0.0000392 per credit. Break-even for Startup: $49 / $0.0000475 = 1,031,579 characters. Scale: $299 / $0.0000475 = 6,294,737.

The 5.5% platform fee from Sume's pricing policy lifts the Sume rate to about $0.050113 per 1,000 characters and pulls the break-evens down to roughly 977,800 and 5,966,575.

When each side is the better fit

• Under about 1 million characters a month, pay per character on Sume and skip the monthly fee.
• Between 1 and 6 million, Cartesia's Startup plan is cheaper per character if you use the allowance every month.
• Above 6 million, Scale is cheaper per character, and the gap grows with volume.
• If you need Cartesia's own agent, telephony or enterprise terms, those are outside what Sume's TTS endpoint sells.

Turning characters into minutes

Cartesia's page says a minute of generated audio takes 750 to 800 credits. At 800 characters a minute, 1 million characters is 1,250 minutes of audio, about 20.8 hours. On Sume that million characters costs 1,000 x $0.0475 = $47.50, just under the $49 Startup fee. That is why the break-even lands close to a million characters, or roughly 21 hours of narration a month.

At the 750-character end of Cartesia's range the same million characters is 1,333 minutes. The dollar figure does not change, because Sume bills by character and not by minute. A 12-minute narration at 800 characters a minute is 9,600 characters, which is $0.456 on Sume, and it must be sent in requests of at most 20,000 characters.

What Sume does not do

Sume does not offer a monthly credit bundle for TTS, does not sell Cartesia's voice agent, and does not include any non-Sonic engine in TTS Router yet. This comparison covers plan fees only; overage rates on Cartesia's plans were not on the page we read, so none are assumed here.