NVIDIAFeb 2026
How Magpie-Multilingual 357M compares with the other 102 text to speech models.
Middle of the pack. Ranked #61 of 103, 270 Elo points behind #1 and ahead of 41% of models.
Close race with 24 models. Its score range overlaps with TTS-1, OpenAudio S1, Gemini 2.5 Flash Lite TTS and 21 more, so you may not notice a difference.
Usually loses. Against a typical model it gets picked in about 47% of blind votes.
Rank
#61
of 103 models
Elo rating
1,064
270 behind #1
typical 1,086
± range
±12
how certain the score is
Votes
–
blind comparisons
This model next to its closest competitors. Click another bar to open that model.
How often people would pick this model over another in a blind vote, based on the Elo scores.
People see two results for the same prompt without knowing which model made them, and vote for the better one. Gaps under about 20 Elo points are hard to notice in practice.
Benchmark data from Artificial Analysis, an independent benchmarking company. Rankings refresh every few hours.
kitio.ai