platform-kit/mars5-tts

A novel speech model for insane prosody.

Capabilities

No capability data available

Community model (estimated from hardware time)

Name	Type	Description	Default	Constraints
`freq_penalty`	integer	—	`3`	—
`ref_audio_file`	string(uri)	Reference audio file to clone from <= 10 seconds	`"https://replicate.delivery/pbxt/L9a6SelzU0B2DIWeNpkNR0CKForWSbkswoUP69L0NLjLswVV/voice_sample.wav"`	—
`ref_audio_transcript`	string	Text in the reference audio file	`"Hi there. I'm your new voice clone. Try your best to upload quality audio."`	—
`rep_penalty_window`	integer	—	`95`	—
`temperature`	number	—	`0.5`	min: 0, max: 5
`text`	string	Text to synthesize	`"Hi there, I'm your new voice clone, powered by Mars5."`	—
`top_k`	integer	—	`95`	min: 0, max: 100

freq_penaltyinteger

Default: 3

ref_audio_filestring

Reference audio file to clone from <= 10 seconds

Default: "https://replicate.delivery/pbxt/L9a6SelzU0B2DIWeNpkNR0CKForWSbkswoUP69L0NLjLswVV/voice_sample.wav"

ref_audio_transcriptstring

Text in the reference audio file

Default: "Hi there. I'm your new voice clone. Try your best to upload quality audio."

rep_penalty_windowinteger

Default: 95

temperaturenumber

Default: 0.5min: 0, max: 5

textstring

Text to synthesize

Default: "Hi there, I'm your new voice clone, powered by Mars5."

top_kinteger

Default: 95min: 0, max: 100

Version: 6aed0f11f3baUpdated: 7/25/2026541 runs