Inputs
Audio for understanding/transcription, or text for synthesis, using the selected model's exact api_mode.
audio-speech capability center
Audio API: understanding and speech-to-speech building blocks
Use verified audio-understanding, ASR, and TTS endpoint modes without treating every speech model as interchangeable.
Contract
Inputs, outputs, and lifecycle.
Outputs
Text analysis/transcript or generated audio; a direct speech-to-speech result is not implied unless catalog metadata says so.
Endpoint
Model-specific: /v1/chat/completions, /v1/audio/transcriptions, or /v1/audio/speech
Lifecycle
The listed audio endpoints are synchronous; a speech-to-speech workflow chains transcription or audio understanding to a separate TTS request.
Supported tasks
One definition per task.
audio-understanding
Limit: A TTS or ASR label does not prove direct audio understanding, voice cloning, diarization, or speech-to-speech support.
Matching catalog
14 models with explicit metadata.
Unknown capabilities are omitted, never guessed from a model name. Pricing examples use the synchronized public data; live Dash Pricing remains authoritative.
stepaudio-2-asr-pro
StepFun StepAudio 2 ASR Pro — 32B-parameter speech recognition, the accuracy-first option in the StepFun ASR line where stepaudio-2.5-asr is the speed-and-cost option. Accepts ogg, mp3, and wav uploads; transcript output is not billed.
Tasks: video-with-audio, transcription
Endpoint: POST /v1/audio/transcriptions
Cost example: 1 audio hour × $0.35 = $0.35.
step-asr
StepFun step-asr — general-purpose speech recognition covering Chinese, English, and Chinese dialects, for transcription, meeting minutes, call quality review, and voice search. Accepts ogg, mp3, and wav uploads; transcript output is not billed.
Tasks: video-with-audio, transcription
Endpoint: POST /v1/audio/transcriptions
Cost example: 1 audio hour × $0.15 = $0.15.
step-asr-1.1
StepFun step-asr-1.1 — the updated general-purpose recognizer in the step-asr line, covering Chinese, English, and Chinese dialects. Accepts ogg, mp3, and wav uploads; transcript output is not billed.
Tasks: video-with-audio, transcription
Endpoint: POST /v1/audio/transcriptions
Cost example: 1 audio hour × $0.35 = $0.35.
mimo-v2.5-asr
Xiaomi MiMo-V2.5-ASR — speech recognition for Chinese, English, code-switching, regional dialects, noisy recordings, far-field audio, multiple speakers, and lyrics.
Tasks: video-with-audio, transcription
Endpoint: POST /v1/chat/completions
Cost example: 1 audio hour × $0.074 = $0.074.
mimo-v2.5-tts
Xiaomi MiMo-V2.5-TTS — expressive multilingual speech synthesis with built-in voices, natural-language style instructions, emotion, pace, tone, dialect, and character control.
Tasks: video-with-audio, text-to-speech
Endpoint: POST /v1/chat/completions
Cost example: 1 10k characters × $0 = $0.
stepaudio-2.5-tts
StepAudio 2.5 TTS — contextual speech synthesis with inline natural-language direction, expressive delivery, built-in voices, and OpenAI-compatible audio output.
Tasks: video-with-audio, text-to-speech
Endpoint: POST /v1/audio/speech
Cost example: 1 10k characters × $0.85 = $0.85.
glm-tts
GLM-TTS — context-aware speech synthesis with expressive delivery, streaming output, fast first audio, and controls for voice, speed, and volume.
Tasks: video-with-audio, text-to-speech
Endpoint: POST /v1/audio/speech
Cost example: 1 10k characters × $0.307692 = $0.307692.
qwen3-asr-flash
Qwen3-ASR-Flash — multilingual file transcription with language hints, inverse text normalization, emotion recognition, and OpenAI-compatible audio input.
Tasks: video-with-audio, transcription
Endpoint: POST /v1/audio/transcriptions
Cost example: 1 audio hour × $0.123539 = $0.123539.
qwen3-tts-flash
Qwen3-TTS-Flash — low-latency multilingual speech synthesis with mixed-language input, built-in voices, automatic language matching, and character-based billing.
Tasks: video-with-audio, text-to-speech
Endpoint: POST /v1/audio/speech
Cost example: 1 10k characters × $0.123077 = $0.123077.
speech-2.8-hd
MiniMax Speech 2.8 HD — high-fidelity speech synthesis with natural emotional delivery, language enhancement, voice controls, and production audio formats.
Tasks: video-with-audio, text-to-speech
Endpoint: POST /v1/audio/speech
Cost example: 1 10k characters × $0.538462 = $0.538462.
speech-2.8-turbo
MiniMax Speech 2.8 Turbo — speed-focused speech synthesis with natural output, emotion controls, language enhancement, and common production audio formats.
Tasks: video-with-audio, text-to-speech
Endpoint: POST /v1/audio/speech
Cost example: 1 10k characters × $0.307692 = $0.307692.
step-tts-2
Step TTS 2 — expressive speech synthesis with official and cloned voices, speed and volume control, pronunciation mapping, and multiple output formats.
Tasks: video-with-audio, text-to-speech
Endpoint: POST /v1/audio/speech
Cost example: 1 10k characters × $0.4 = $0.4.
step-tts-mini
Step TTS Mini — cost-efficient expressive speech synthesis for Chinese, English, Japanese, Cantonese, and Sichuanese, with official and cloned voices.
Tasks: video-with-audio, text-to-speech
Endpoint: POST /v1/audio/speech
Cost example: 1 10k characters × $0.15 = $0.15.
stepaudio-2.5-asr
StepAudio 2.5 ASR — a 4B MTP streaming transcription model for fast Chinese-English recognition, ITN normalization, subtitles, meetings, and voice agents.
Tasks: video-with-audio, transcription
Endpoint: POST /v1/audio/transcriptions
Cost example: 1 audio hour × $0.022 = $0.022.
Run or estimate before registration.
Calculate API cost · Generate curl, Python, or JavaScript · Install the Seedance 2 Agent Recipe
Source and method: public model pricing dataset, updated 2026-08-08. Limitations are stated above; third-party claims remain next to the model pages that cite them.