Audio Models

Model	Context	Max output	Official price	Verified price	Discount
speech-2.8-hdText to speech	10K	—	Input$100.00Output$0.00/1M characters	—	—
cosyvoice-v3.5-plusText to speech	—	—	Input$22.06Output$0.00/1M characters	—	—
eleven-v4-turboText to speech	—	—	Input$40.00Output$0.00/1M characters	—	—
fish-transcribe-1Speech to Text	—	—	$0.0001/sec	—	—
fun-asrSpeech to Text	—	—	$0.000035/sec	—	—
fun-asr-flash-2026-06-15Speech to Text	—	—	$0.000035/sec	—	—
grok-sttSpeech to Text	—	—	$0.00002778/sec	—	—
grok-ttsText to speech	—	—	Input$15.00Output$0.00/1M characters	—	—
grok-voice-latest	—	—	$0.001333/sec	—	—
grok-voice-sttSpeech to Text	—	—	$0.00002778/sec	—	—
grok-voice-think-fast-2.0	—	—	$0.001333/sec	—	—
grok-voice-ttsText to speech	—	—	Input$15.00Output$0.00/1M characters	—	—
kling-ttsText to speech	—	—	$0.007/req	—	—
minimax-music-2.6Music	—	—	$0.15/req	—	—
minimax-speech-2-8Text to speech	—	—	Input$100.00Output$0.00/1M characters	—	—
paraformer-8k-v2Speech to Text	—	—	$0.00001177/sec	—	—
paraformer-v2Speech to Text	—	—	$0.00001177/sec	—	—
qwen3-asr-flashSpeech to Text	—	—	$0.000035/sec	—	—
qwen3-asr-flash-filetransSpeech to Text	—	—	$0.000035/sec	—	—
qwen3-tts-1-7b-customvoiceText to speech	—	—	Input$15.00Output$0.00/1M characters	—	—
qwen3-tts-1-7b-voicedesignText to speech	—	—	Input$15.00Output$0.00/1M characters	—	—
qwen3-tts-flashText to speech	—	—	Input$10.00Output$0.00/1M characters	—	—
speech-2.8-turboFastText to speech	—	—	Input$60.00Output$0.00/1M characters	—	—
stable-audio-3Music	—	—	$0.26/req	—	—

1 / 2

All model series

6 series

Kling

Kuaishou

ImageVideoAudioText to speech+2

Auto pricing units vary by model

15 modelsOpen series

MiniMax

ChatImageVideoAudioText to speechMusic+4

Auto pricing units vary by model

11 modelsOpen series

Grok Voice

xAI

AudioText to speechSpeech to Text+1

Auto pricing units vary by model

6 modelsOpen series

Qwen Omni

Alibaba Cloud

ChatAudioTranslationSpeech to TextVision+3

Auto from $0.15 · per 1M tokens

4 modelsOpen series

Stable Audio

Stability AI

AudioMusic

Auto from $0.0206 · per request

4 modelsOpen series

Qwen TTS

Alibaba Cloud

AudioText to speechMusic+1

Auto from $10.00 · per 1M characters

3 modelsOpen series

Match the audio API to the task

Speech generation, transcription, and real-time conversation have different inputs, outputs, and latency requirements. An audio category label does not mean every model supports all three.

Selection signals

Choose the required operation first, then verify accepted audio formats, languages, and length limits.
Compare the relevant units, which may include audio duration, text characters, or tokens.
Evaluate accents, background noise, pronunciation, and turn-taking on representative recordings.

Docs Model rankings Compare models

All model series

Kling

MiniMax

Grok Voice

Qwen Omni

Stable Audio

Qwen TTS

Match the audio API to the task

Selection signals

Related pages