Cartesia Pro $5 for 100K credits: 133 minutes vs Sume TTS price

Cartesia's pricing page (read 2026-10-02) lists Pro at $5 a month for 100K credits, which it translates to about 133 minutes of Sonic-3.6 text-to-speech. That is roughly $0.038 per minute of speech on Pro, falling to about $0.028 per minute on Scale. Sume's text-to-speech is billed per character, so you compare by converting your own script length.

Cost per minute by plan

The page prints monthly fee, credits and approximate TTS minutes per plan, and notes that it does not state a credit cost per character. The per-minute figure below is the fee divided by the listed minutes, so it assumes you use the whole allowance.

Sume's side: characters, not minutes

Sume's text-to-speech is priced at $0.0475 per 1,000 characters on the API pricing page, spaces and punctuation included, with a 20,000-character maximum per request. Its TTS Router lists Cartesia's Sonic ids, so the voice engine is the same family; Sume's own documentation says the price is Cartesia's list rate times 1.25.

To compare, you need characters per minute for your voice and script. These are illustrative inputs, not measurements: a script that runs 500 characters per minute costs $0.024 per minute on Sume, 1,000 characters per minute costs $0.0475, and 1,500 costs about $0.071. Replace them with your own.

What each side is good for

Cartesia's plans are a good fit if you talk to Cartesia directly, use its voice agents or streaming, and will use most of an allowance every month. Sume's per-character price suits irregular use: no monthly fee, one wallet shared with video and image jobs, and one usage ledger to reconcile.

Neither number is the whole bill. A script is rarely right the first time, so budget for retakes, and measure your first render: read the characters Sume billed and divide by the audio minutes you got.