Step TTS 2

Name: Step TTS 2
Author: StepFun

ttsproprietary

End-to-end TTS with 11 emotions, 17 styles, 3 languages (Mandarin, Cantonese, Sichuan dialect, Japanese). 10s audio voice cloning.

Capabilities

Show unsupported (13)

No supported capabilities recorded.

Details

Model IDstep-tts-2

ProviderStepFun

Creatorstepfun

Familystep-tts

Licenseproprietary

Statusactive

Typetts

Input modalitiesaudio, text

Output modalitiesspeech

Open weightNo

Parameters—

Architecture—

Max chars—

Sample rate—

Voices—

Languages—

Release date—

Sourceofficial

Endpoints

audio speechPOST

https://api.stepfun.com/audio-speech

Pricing

2.8 元／万字符 (CNY)

Family Comparison: step-tts

Model	Max chars	Sample rate	Voices	Languages
Step TTS 2	—	—	—	—
Step TTS Mini	—	—	—	—

API

GET/v1/models/stepfun/step-tts-2

Step TTS 2

step-tts-2

End-to-end TTS with 11 emotions, 17 styles, 3 languages (Mandarin, Cantonese, Sichuan dialect, Japanese). 10s audio voice cloning.

Changes · 5 entries

step-tts-2updateca7a400Nov 2, 2026, 03:02 AM

pricing_notes["0.40 USD per 10k characters."]→["2.8 元／万字符 (CNY)"]

taglineEnd-to-end TTS with 11 emotions, 17 styles, 3 languages (...

step-tts-2updateb2f45bfOct 29, 2026, 03:29 AM

statusdeprecated→active

pricing_notes["2.8 元/万字符 (CNY)."]→["2.8 元／万字符 (CNY)"]

step-tts-2update40decdeSep 16, 2026, 06:38 AM

statusactive→deprecated

step-tts-2updatec67ed84Jul 31, 2026, 10:28 AM

pricing_notes["0.40 USD per 10k characters."]→["2.8 元/万字符 (CNY)."]

taglineEnd-to-end TTS with 11 emotions, 17 styles, 3 languages (...

step-tts-2create3e56cf1Apr 26, 2026, 07:23 PM