StepAudio 2.5 TTS (Aug 2026) text-to-speech model
StepFunCNReleased Aug 2026.
StepAudio 2.5 TTS (Aug 2026) is a text-to-speech (TTS) model from StepFun. It ranks #15 of 98 on the Artificial Analysis TTS arena with Elo 1214 (95% CI ±14, 2,254 samples). List price is about $85 per 1M characters. Released Aug 2026.
Key facts
- Provider
- StepFun
- AA rank
- #15 / 98
- AA Elo (95% CI)
- 1214 ±14
- AA samples
- 2,254
- TTS Arena Elo (uncertainty)
- –
- TTS Arena votes
- –
- TTS Arena win rate
- –
- Price (per 1M chars)
- $85
- Released
- Aug 2026
- Open weights
- –
- License
- –
- Voices tested by AA
- 8
- Languages
- –
- Voice cloning
- –
"–" means the upstream sources do not provide this value; we do not guess.
Nearby models
By AA Elo; this model highlighted| # | Model | AA Elo | Arena Elo | Price $/1M chars | Open weights |
|---|---|---|---|---|---|
| 12 | IRealtime TTS-2 Flash Inworld | 1219 | – | $10.4 | – |
| 13 |  Voxtral 3 TTS Mistral | 1219 | – | $13 | – |
| 14 |  v3 Conversational ElevenLabs | 1215 | – | $50 | – |
| 15 | StepAudio 2.5 TTS (Aug 2026) StepFun | 1214 | – | $85 | – |
| 16 |  Gemini 3.1 Flash TTS Google | 1213 | – | $18.3 | – |
| 17 | VLuna TTS VUI Labs | 1210 | – | $15 | – |
| 18 | CSonic 3.5 Cartesia | 1192 | – | $49 | – |
Alternatives · Cheaper with AA Elo ≥ 1184
| # | Model | AA Elo | Arena Elo | Price $/1M chars | Open weights |
|---|---|---|---|---|---|
| 1 | Eleven v4 Turbo ElevenLabs | 1336 | – | $40 | – |
| 2 | Eleven v4 ElevenLabs | 1332 | – | $80 | – |
| 3 |  Qwen-Audio-3.1-TTS-Plus Alibaba | 1307 | – | $19.3 | – |
| 4 | CSonic 3.6 Cartesia | 1292 | – | $49 | – |
| 5 | Gemini 3.8 Flash TTS Google | 1280 | – | $16.5 | – |
Alternatives · More from StepFun
| # | Model | AA Elo | Arena Elo | Price $/1M chars | Open weights |
|---|---|---|---|---|---|
| 32 | Step TTS 2 (Mar 2026) StepFun | 1140 | – | $40 | – |
| 41 | Step Audio EditX (Mar 2026) StepFun | 1100 | – | – | Open |
Alternatives · Top 5 open-weight by AA Elo
| # | Model | AA Elo | Arena Elo | Price $/1M chars | Open weights |
|---|---|---|---|---|---|
| 9 |  Fish Audio Drama 3 Preview Fish Audio | 1248 | – | $15 | Open |
| 11 | BBreeze TTS 2 BreezeBlue | 1230 | – | $34 | Open |
| 19 |  Ming-3.0-tts Ant Group | 1191 | – | – | Open |
| 38 | Fish Audio S2 Pro Fish Audio | 1110 | – | $15 | Open |
| 41 | Step Audio EditX (Mar 2026) StepFun | 1100 | – | – | Open |
FAQ
Where does StepAudio 2.5 TTS (Aug 2026) rank among TTS models?
As of Oct 25, 2026, StepAudio 2.5 TTS (Aug 2026) ranks #15 of 98 on the Artificial Analysis TTS leaderboard with Elo 1214 (±14).
Is StepAudio 2.5 TTS (Aug 2026) open source?
The sources do not list StepAudio 2.5 TTS (Aug 2026) as open-weight; it is generally used through the StepFun API.
What is the best Chinese TTS model?
By AA Elo, the top model from a Chinese provider is Qwen-Audio-3.1-TTS-Plus by Alibaba (Elo 1307, #3 overall).
Which top-10 TTS model is cheapest?
Among the AA top 10, the lowest list price is Simba 3.2 (SpeechifyAI) at $6.6 per 1M characters, Elo 1246.
What is the best open-source TTS model?
By AA Elo, the top open-weight model is Fish Audio Drama 3 Preview by Fish Audio (Elo 1248, #9 overall); 18 open-weight models have an AA score.
Sources
"–" means the upstream sources do not provide this value; we do not guess.