Gemini 3.8 Flash TTS vs Sume TTS: break-even 284 chars a minute

Gemini 3.8 Flash TTS bills the audio it produces and Sume TTS bills the characters you send, so the break-even is a speaking rate: 284.2 characters per minute of finished audio on Flash at the current $9.00 per million audio tokens, rising to 568.4 characters per minute from 2027 when the listed price doubles. Sume is cheaper than Gemini's audio-output charge only when your narration runs slower than that rate. Normal read-aloud copy is far above 284 characters a minute, so on this single measure Gemini is cheaper at the current price.

How the break-even is computed

Google's pricing page says audio tokens correspond to 25 tokens per second of audio. One minute is 60 x 25 = 1,500 tokens. At $9.00 per million tokens that is 1,500 x 9 / 1,000,000 = $0.0135 per minute, which matches the page's own $0.00225 per 10 seconds. Sume's rate is $0.0475 per 1,000 characters, or $0.0000475 per character. Dividing the Gemini minute cost by the Sume character cost gives the number of characters per minute at which the two bills are equal.

This ignores Gemini's text-input charge ($0.50 per million text tokens to the end of 2026), which makes Gemini slightly dearer than the table shows, and it ignores Sume's 1-cent job floor, which makes Sume dearer for short jobs. Gemini's token count for a given text is also not the same as a character count, so do not read the table as a quote.

Why the two meters behave differently

A seconds-based meter charges for pauses, slow delivery and padded silence. A character-based meter does not: the same 900-character script costs the same on Sume at any pace, and you can set a speed on the job. At 900 characters a minute, a 10-minute read is 9,000 characters, or $0.4275 on Sume against 10 x $0.0135 = $0.135 on Gemini Flash audio output, before Gemini's input tokens. Slow the Gemini read by 20 percent and its audio bill rises with the extra seconds while the Sume bill does not move.

The reverse also holds: a script that is mostly short words read quickly is cheap on Gemini and dear on Sume. Measure your own characters per minute from a finished read and plug it into the table before choosing.

What to check before switching

Gemini 3.8 Flash TTS and Flash-Lite TTS are listed on Google's pricing page, but Sume's catalog lists only Sume TTS 1.0 and the TTS Router, so Gemini TTS is a separate integration whose output you would have to import to use in a Sume Timeline (Timeline audio states all URLs must be media.sume.com audio). The price difference only matters if the voice step is a large share of your spend; for one-off ad reads it is cents.