StepFun

Model ProviderOpenAI compatible

Chinese AI startup founded by ex-Microsoft engineers, building the Step series of foundation, reasoning, vision, and audio models.

Active models

25+ 13 deprecated

Families

Founded2023

API Base URL

Type

Capabilities

OSSShow deprecated (13)

Model	Caps	Type	Pricing
Step 3.5 Flash 2603OSSstep-3.5-flash-2603	RTJJMSF	reasoning	$0.1$0.29
Step 3.5 FlashOSSstep-3.5-flash	RTJJMSF	reasoning	$0.1$0.29
Step-1o Turbo Visionstep-1o-turbo-vision	TS	chat	$0.35$1.11
Step-1o Audiostep-1o-audio	S	audio	$3.47$8.33
StepAudio 3 ASR Maxstepaudio-3-asr-max	—	transcription
StepAudio 3 Chatstepaudio-3-chat-preview	—	audio
StepAudio 3 Genstepaudio-3-gen-preview	—	audio
StepAudio 3 Musicstepaudio-3-music-preview	—	audio
StepAudio 3 Realtimestepaudio-3-realtime-preview	—	audio
StepAudio 3 TTSstepaudio-3-tts	—	tts

1–10 of 25

1 / 3