whatmodel.sh / tts / Deepgram Aura 2

Deepgram Aura 2

+ contribute
Update model informationAdd a place to run itAdd a test resultPropose a benchmarkCorrect a benchmark scoreEdit data on GitHub

Deepgram Aura 2 is a closed-weights text-to-speech model from Deepgram, released April 2025.

Catalogue identity notes

DeepGram Aura-2 as explicitly linked in the result row.

Model information sources

Release date ↗

56.5±3.1#50 of 118 text-to-speech modelsfrom 6 benchmark results across 4 sources

Score within the Vapi Humanness · Bradley–Terry rating, 2026-09-25 snapshot comparison set only.

$ how it compares

The text-to-speech models ranked either side of Deepgram Aura 2, out of 148 with a score.

#modelscorewhereRAMreleased
60F5-TTS Base (v0) SWivid (SJTU X-LANCE)57±2local~1.4GB2024-10
61TTS-1 HD OpenAI57±7api $30.00/M chars2023-11
62Deepgram Aura 2 Deepgram57±3api $30.00/M chars2025-04
63ZONOS2 8B Zyphra57±2local~5.2GB2026-06
64Supertonic 3 Supertone55±10local~0.4GB2026-05

see all text-to-speech models · find the ones that fit your machine

$ what it's good at

other tests6 not in a task · 96% of its score
testresultvs the others in its rundifficultycountswhy
EmergentTTS · 6-category preference25.8% ✎▅▅▃▃█▅▅▃·▃·▅ beats 7/16hard · avg 30%▇▇▇▇▇▇▇▇▇▇▇▇ 30%at its level
Rime ContentBench · paired human preference vs Rime Coda (US English)0.15 / 2 ✎▄·▄········█ beats 2/3medium · avg -0.58 / 2▇▇▇▇▇▇▇▇ 20%at its level
Rime SupportBench · paired human preference vs Rime Coda (US English)-0.37 / 2 ✎█·······█·██ beats 1/3hard · avg -0.90 / 2▇▇▇▇▇▇▇ 18%at its level
RW-Voice-EQ TTS · Reliability (human)3.60 / 5 ✎█·▄······▄·▄ beats 1/4medium · avg 3.01 / 5▇▇▇▇▇ 13%at its level
EmergentTTS · v1 paper automatic MOS (WV-MOS)4.33 / 5 ✎▄·█·█▄····█▄ beats 8/8easy · avg 3.93 / 5▇▇▇ 7.3%blunt
EmergentTTS · 6-category WER16.8% ✎▂▄··▂▄▄··█▄▆ beats 7/16avg 16%▇▇▇ 6.9%blunt

$ where to run it

Through an API

api · $30.00/M chars · deepgram.com ↗ · price sourceUSD per 1M input characters: Deepgram lists Aura-2 text-to-speech at $0.030 per 1k characters, pay as you go (x1000); Growth plan $0.027 per 1k. Product: Deepgram TTS API. API model: aura-2. Settings: pay-as-you-go. Price basis: published list rate. Exclusions: Growth-plan prepay discount. Currency: USD. Deepgram official pricing page, observed 2026-09-25.

$ how the score was worked out · Rime Alpha TTS Benchmark v1.0… 38%, EmergentTTS-Eval six-category… 37%, RW-Voice-EQ Bench (Hume AI) 13% and 1 more

Each result implies a score on its own, after adjusting for how hard its test is and who ran it. The score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its counts column above; here it's added up by who ran the tests. How it works

sourceresultsshare of the scorelicence
Rime Alpha TTS Benchmark v1.0.0 · SupportBench and ContentBench paired human preference2▇▇▇▇▇▇▇▇▇▇▇▇ 38%CC BY 4.0
EmergentTTS-Eval six-category leaderboard2▇▇▇▇▇▇▇▇▇▇▇ 37%Apache-2.0
RW-Voice-EQ Bench (Hume AI)1▇▇▇▇ 13%CC BY 4.0
EmergentTTS-Eval v1 paper (arXiv:2505.23009v1) · Table 1 automatic MOS1▇▇ 7%CC BY 4.0
prior · pull toward the middle▇ 4%
weighted average= 56.5

$ improve this page

Update model informationAdd a place to run itAdd a test resultPropose a benchmarkCorrect a benchmark scoreEdit data on GitHub

Use a short GitHub form or edit the data through a pull request. Every number needs a source; here's which ones we can use.

data updated · sources and licences · json