Skip to content

Comparison · read on October 27, 2026

Gradium TTS vs Smallest.ai Lightning v3.1

Two text-to-speech (TTS) models side by side, from four public benchmarks. Speed comes first, then listener rating and price; the three are never added into a score.

MeasureGradium TTSSmallest.ai Lightning v3.1
ProviderGradiumSmallest.ai
Speed tierHeadroomNot timed
P95 time to first audio104.3 msNot timed
Median time to first audio53.7 msNot timed
Listener Elo1,1741,020
Price per 1M characters$36 to $57.8$17.5 to $25
AccessProprietaryProprietary
Rank in the full ranking271
Rank by use caseCustomer service 1, Voice agents 2, Assistants 1, Entertainment 1, Knowledge sharing 1Customer service 47, Assistants 58, Entertainment 58, Knowledge sharing 58

Which should you choose?

Where a model is not timed or not rated, nothing is estimated. See each model page for its sources: Gradium TTS and Smallest.ai Lightning v3.1.

Frequently asked questions

Which is faster, Gradium TTS or Smallest.ai Lightning v3.1?
Speed: not compared, because a benchmark in this index does not publish it for both models.
Which is rated higher by listeners, Gradium TTS or Smallest.ai Lightning v3.1?
Listener rating: Gradium TTS comes out ahead (Elo 1,174 against 1,020; scores within each other's 95% intervals are a tie).
Which is cheaper, Gradium TTS or Smallest.ai Lightning v3.1?
Price: Smallest.ai Lightning v3.1 comes out ahead ($36 to $57.8 against $17.5 to $25 per million characters, lowest list price).
Why is this model always on the left?
Gradium TTS is the reference model of this site's comparison pages: it is set against every other listed model. The editor chose it; it is not chosen by rank.