Skip to main content
Find the right model for voice agents, translation, transcription, speech generation, and music creation. This page summarizes what each model does; see the individual model pages for specs and quick-start code. Try them in Voice Studio. Language support:
  • TTS and ASR — Chinese, English, Japanese, Korean, French, Spanish
  • Realtime — Chinese, English
Note: All languages except Chinese and English are in preview. More languages are on the way.

​
Choose a model

​
StepAudio 3 TTS

Generate natural, expressive speech—including laughter, hesitation, and self-correction—with low-latency streaming and fine-grained control over context. Learn more about StepAudio 3 TTS.

​
StepAudio 3 ASR

Context-aware transcription for proper names, specialized terminology, dialects, Chinese–English code-switching, and challenging audio such as whispers and singing. Learn more about StepAudio 3 ASR.

​
StepAudio 3 Realtime

Full-duplex real-time voice: build natural, real-time voice agents that handle interruptions, take turns smoothly, reason while speaking, and call tools. Learn more about StepAudio 3 Realtime.

​
StepAudio 3 Gen

Create speech, singing, sound effects, ambience, and background music in a single clip from a natural-language prompt. Learn more about StepAudio 3 Gen.

​
StepAudio 3 Music

Turn lyrics, style prompts, or dry vocals into complete songs, instrumentals, covers, and soundtracks. Learn more about StepAudio 3 Music.

​
StepAudio 2.5 Chat

An end-to-end speech-understanding model served through an OpenAI-compatible Chat Completion API. It accepts audio or text input and returns a text response, interpreting not just the words but also vocal cues such as intonation, hesitation, and laughter.
  • Paralinguistic understanding: captures a speaker’s emotional state and intent from vocal cues a text transcript would miss.
  • Unified text and audio input: typed and spoken turns share the same endpoint, so one conversation can mix both.
  • For spoken replies, pass the text output to a text-to-speech model such as StepAudio 2.5 TTS.
Common use cases: voice assistants and conversational agents, call and meeting analysis, and voice message triage.

​
StepAudio 2.5 Realtime

An end-to-end speech-to-speech model served over a WebSocket Realtime API. It takes streaming audio or text and responds with streaming audio and a matching text transcript—no separate transcription or synthesis step—keeping latency low enough for natural back-and-forth conversation.
  • Paralinguistic understanding: interprets vocal cues such as intonation, pacing, and hesitation in addition to the spoken words.
  • Natural turn-taking: server-side voice activity detection (VAD) lets the model reply at the right moment and be interrupted mid-response.
  • System voices: a set of English system voices selectable per session; see the voices list.
Common use cases: voice agents for customer support, live hands-free assistants, and interactive phone and IVR systems.

​
StepAudio 2.5 TTS

A text-to-speech model that brings context understanding into the entire generation pipeline. Instead of matching preset tags, you describe the delivery you want in natural language, at two levels:
  • Global context sets the overall mood, scene, and character relationships for an entire passage.
  • Inline context fine-tunes how individual words and phrases are delivered.
It supports layered, mixed delivery intents, such as “restrained sadness, no sobbing, with a slight tremble”. The model renders pauses, stress, rhythm, and tonal transitions accordingly, producing speech with natural breathing, dynamic emphasis, and emotional arcs. Zero-shot voice cloning: about 3 seconds of reference audio is enough to clone a voice precisely, and cloned voices keep full global and inline context control, with no fixed voice library or preset characters required. Common use cases: audiobooks, dubbing, ad narration, emotional storytelling, and other content that demands high vocal expressiveness.

​
StepAudio 2.5 ASR

A new-generation speech recognition model built on a 4B-parameter Multi-Token Prediction (MTP) architecture, maintaining SOTA transcription accuracy while sharply reducing latency. Supports Chinese and English recognition with ITN text normalization. It offers two access methods:
  • One-shot recognition (stepaudio-2.5-asr): submit audio over HTTP + SSE and receive the transcription streamed back incrementally.
  • Real-time bidirectional streaming (stepaudio-2.5-asr-stream): stream audio over WebSocket and receive incremental and final results with per-word timing; supports server-side VAD.
  • Suitable for live captions, voice input, meeting transcription, and back-end batch processing.

​
Usage limits

  • Text-to-speech input: up to 1,000 characters per request.
  • Text-to-speech output formats: wav, mp3, flac, opus, pcm; the default is mp3.
  • StepAudio 3 Gen input limits: the instruction and roles fields are capped at 500 characters and scripts at 1,000 characters; exceeding a limit returns HTTP 400.
  • StepAudio 3 Gen delivery: currently supports HTTP audio / sse return; bidirectional WebSocket streaming is not supported.

​
Next steps

StepAudio 3 TTS

Human-level, low-latency text-to-speech.

StepAudio 3 ASR

Context-aware recognition on a large language model foundation.

StepAudio 3 Realtime

Full-duplex real-time voice with adaptive reasoning and Voice Agent.

StepAudio 3 Gen

Unified audio generation—voice, SFX, ambience, and music.

StepAudio 3 Music

Generate songs, instrumentals, covers, and scoring from dry vocals.

StepAudio 2.5 Chat Overview

Explore the end-to-end speech-understanding model with audio and text input.

StepAudio 2.5 Realtime Overview

Explore the end-to-end speech-to-speech model with real-time conversation over WebSocket.

StepAudio 2.5 TTS Overview

Explore the contextual TTS model with dual-level context control and zero-shot voice cloning.

StepAudio 2.5 ASR Overview

Explore the new-generation 4B MTP ASR model with one-shot SSE and real-time bidirectional streaming.

Voice interaction developer guide

Get started with speech generation, voice cloning, and automatic speech recognition.