CekuraBench

Cartesia Sonic 3.5

Every voiceCompare

8th of 26 voices at starting to talk fast

Non-streaming latency (median): 261 ms

Cartesia Sonic 3.5: 8th of 26 voices at starting to talk fast. Its weakest is read from context, 26th of 26.

Last updated Methodology by Dileep Chagam

Cartesia Sonic 3.5 at a glance

Where it ranks among all 26 voices, in plain words.

Good at

Nothing in the top quarter of the voices.

Not so good at
  • Keeps talking after a stop6th of 6
  • Misreads read from context26th of 26
  • Misreads numbers and shorthandTied 25th of 26
  • Misreads hard parts24th of 26
  • Misreads codes and addressesTied 20th of 26

Hear Cartesia Sonic 3.5

The first read of each test sentence, unedited.

Cartesia Sonic 3.5

Cartesia Sonic 3.5 head to head

Pick any voice to set against it. Every pair has its own page.

Compare
MeasureCartesia Sonic 3.5Soniox TTS RT v2
  1. Non-streaming261 ms241 ms
  2. Streaming244 ms183 ms
  3. Stopping809 ms0 ms
  4. Accuracy73.1%83.6%
  5. Naturalness4.414.37
  6. Price / 1M$39.20$14.81
  7. Measures won1 of 65 of 6
The check marks the better value on each measure. Every pair has its own page with the full numbers.Open this pair · 325 pairs in all

Its closest rivals on speed

Cartesia Sonic 3.5 FAQ

  • Cartesia Sonic 3.5 starts in 261 ms on a typical sentence (median) and 296 ms on the slowest 1 in 10 (p90), 8th of 26 voices.

  • Yes. With streamed text, Cartesia Sonic 3.5 starts 244 ms after the last word arrives.

  • Yes. After the stop it still sends 809 ms of audio (median), which the agent throws away.

  • Cartesia Sonic 3.5 read 73.1% of the hard parts right, as speech-to-text heard it. Its weakest kind is read from context, at 54.2%. This is an estimate, not a listening test.

  • Cartesia Sonic 3.5 lists at $39.20 per million characters.