Speech-to-speech APIs and models compared
74 speech-to-speech models from 21 providers, with languages, features, prices and shutdown dates.
The cheapest published USD price is Speech Realtime DeepSeek V4 Flash from Yandex SpeechKit at $0.78 per hour; 13 of 74 have a comparable price.
Of these, 3 are open weights, 58 stream over WebSocket or WebRTC and 10 have an announced shutdown date.
Prices are normalized: USD per hour of audio, cheapest first. Tick models to compare them side by side, or refine these filters.
| Compare | Model | Price ↑ |
|---|---|---|
speech-realtime-deepseek-v4-flashSpeech Realtime DeepSeek V4 FlashYandex SpeechKit, S2S, alpha |
$0.78 per hour | |
speech-realtime-max-tts-liveSpeech Realtime Max TTS LiveYandex SpeechKit, S2S |
$0.78 per hour | |
gpt-realtime-translateGPT-Realtime-TranslateOpenAI, S2S |
$2.04 per hour | |
ultravox-v0.6Ultravox v0.6 (Llama 3.3 70B)Inworld AI, S2S |
$3.00 per hour | |
ultravox-v0.6-gemma3-27bUltravox v0.6 (Gemma 3 27B)Inworld AI, S2S |
$3.00 per hour | |
ultravox-v0.7Ultravox v0.7 (GLM 4.6)Inworld AI, S2S |
$3.00 per hour | |
gpt-live-1GPT-Live 1OpenAI, S2S |
$3.00 per hour | |
phonic-1Phonic-1Phonic, S2S |
$4.20 per hour | |
grok-voice-3.0Grok Voice 3.0xAI, S2S |
$4.80 per hour | |
grok-voice-latestGrok Voice (latest)xAI, S2S |
$4.80 per hour | |
eleven_english_sts_v2Eleven English v2ElevenLabs, S2S |
$7.20 per hour | |
eleven_multilingual_sts_v2Eleven Multilingual v2ElevenLabs, S2S |
$7.20 per hour | |
marketplace-stsRespeecher Voice Marketplace Speech-to-SpeechRespeecher, S2S |
$60.00 per hour | |
qwen-audio-3.0-realtime-flashQwen-Audio 3.0 Realtime FlashAlibaba Cloud Model Studio, S2S |
$0.23 per 1M tokens | |
qwen-audio-3.0-realtime-plusQwen-Audio 3.0 Realtime PlusAlibaba Cloud Model Studio, S2S |
$0.80 per 1M tokens | |
qwen-audio-3.1-realtime-flashQwen-Audio 3.1 Realtime FlashAlibaba Cloud Model Studio, S2S |
$0.23 per 1M tokens | |
qwen-audio-3.1-realtime-plusQwen-Audio 3.1 Realtime PlusAlibaba Cloud Model Studio, S2S |
$0.80 per 1M tokens | |
qwen3-omni-flash-realtime-2025-09-15Qwen3-Omni Flash Realtime (2025-09-15)Alibaba Cloud Model Studio, S2S |
$0.52 per 1M tokens | |
qwen3.5-livetranslate-flash-realtimeQwen3.5-LiveTranslate Flash RealtimeAlibaba Cloud Model Studio, S2S |
$7.50 per 1M tokens | |
qwen3.5-livetranslate-flash-realtime-2026-05-19Qwen3.5-LiveTranslate Flash Realtime (2026-05-19)Alibaba Cloud Model Studio, S2S |
$7.50 per 1M tokens | |
qwen3.5-omni-flash-realtimeQwen3.5-Omni Flash RealtimeAlibaba Cloud Model Studio, S2S |
$0.55 per 1M tokens | |
qwen3.5-omni-flash-realtime-2026-03-15Qwen3.5-Omni Flash Realtime (2026-03-15)Alibaba Cloud Model Studio, S2S |
$0.55 per 1M tokens | |
qwen3.5-omni-plus-realtimeQwen3.5-Omni Plus RealtimeAlibaba Cloud Model Studio, S2S |
$2.10 per 1M tokens | |
qwen3.5-omni-plus-realtime-2026-03-15Qwen3.5-Omni Plus Realtime (2026-03-15)Alibaba Cloud Model Studio, S2S |
$2.10 per 1M tokens | |
qwen3.8-livetranslate-flash-realtimeQwen3.8-LiveTranslate Flash RealtimeAlibaba Cloud Model Studio, S2S |
$7.50 per 1M tokens | |
qwen3.8-omni-flash-realtimeQwen3.8-Omni Flash RealtimeAlibaba Cloud Model Studio, S2S |
$0.23 per 1M tokens | |
amazon.nova-2-5-sonicNova 2.5 SonicAmazon Nova Sonic, S2S |
$3.00 per 1M tokens | |
amazon.nova-2-sonic-v1:0Nova 2 SonicAmazon Nova Sonic, S2S |
$3.00 per 1M tokens | |
voice-live-liteVoice Live LiteAzure Speech, S2S |
$15.00 per 1M tokens | |
voice-live-proVoice Live ProAzure Speech, S2S |
$17.00 per 1M tokens | |
voice-live-standardVoice Live StandardAzure Speech, S2S |
$15.00 per 1M tokens | |
audio-mini-realtime-farSpeech-Language LLM Lite (far-field)Baidu AI Cloud Speech, S2S |
¥16.00 per 1M tokens | |
audio-mini-realtime-nearSpeech-Language LLM Lite (near-field)Baidu AI Cloud Speech, S2S |
¥16.00 per 1M tokens | |
audio-realtime-farSpeech-Language LLM Pro (far-field)Baidu AI Cloud Speech, S2S |
¥64.00 per 1M tokens | |
audio-realtime-nearSpeech-Language LLM Pro (near-field)Baidu AI Cloud Speech, S2S |
¥64.00 per 1M tokens | |
qianfan-realtime-flex-v1Speech-Language LLM FlexBaidu AI Cloud Speech, S2S, beta |
– | |
higgs-realtimeHiggs RealtimeBoson AI, S2S |
$0.75 per 1M tokens | |
eleven_realtime_v1_maxEleven Realtime v1 MaxElevenLabs, S2S |
– | |
eleven_realtime_v1_miniEleven Realtime v1 MiniElevenLabs, S2S |
– | |
gemini-2.5-flash-native-audio-preview-12-2025Gemini 2.5 Flash Native Audio Preview (12-2025)Google Cloud, S2S, ends 2027-03-31 |
$3.00 per 1M tokens | |
gemini-3.1-flash-live-previewGemini 3.1 Flash Live PreviewGoogle Cloud, S2S, beta |
$3.00 per 1M tokens | |
gemini-3.5-live-translate-previewGemini 3.5 Live Translate PreviewGoogle Cloud, S2S, beta |
$3.50 per 1M tokens | |
gemini-3.8-liveGemini 3.8 LiveGoogle Cloud, S2S |
$3.00 per 1M tokens | |
gemini-3.8-live-extended-thinkingGemini 3.8 Live Extended ThinkingGoogle Cloud, S2S |
$3.00 per 1M tokens | |
gemini-live-2.5-flash-native-audioGemini Live 2.5 Flash Native Audio (Vertex AI)Google Cloud, S2S, ends 2026-12-13 |
– | |
s2s-translateGradium Speech-to-Speech TranslateGradium, S2S |
– | |
bhashik-stsBhashik Speech-to-Speech TranslationKrutrim, S2S |
– | |
fluid-1Kugel Fluid 1KugelAudio, S2S, beta |
– | |
LiquidAI/LFM2.5-Audio-1.5B-JPLFM2.5-Audio-1.5B-JPLiquid AI, S2S |
– | |
lfm3-audio-2bLFM3-Audio-2BLiquid AI, S2S |
$2.00 per 1M tokens | |
lfm3-omni-2bLFM3-Omni-2BLiquid AI, S2S |
$2.00 per 1M tokens | |
nvidia/nemotron-voicechatNemotron 3 VoiceChatNVIDIA, S2S, beta |
– | |
gpt-audioGPT-AudioOpenAI, S2S, ends 2027-01-20 |
$32.00 per 1M tokens | |
gpt-audio-1.5GPT-Audio-1.5OpenAI, S2S |
$32.00 per 1M tokens | |
gpt-audio-2025-08-28GPT-Audio (2025-08-28)OpenAI, S2S, ends 2027-01-20 |
$32.00 per 1M tokens | |
gpt-audio-miniGPT-Audio miniOpenAI, S2S, ends 2027-01-20 |
$10.00 per 1M tokens | |
gpt-audio-mini-2025-12-15GPT-Audio mini (2025-12-15)OpenAI, S2S, ends 2027-01-20 |
$10.00 per 1M tokens | |
gpt-realtimeGPT-RealtimeOpenAI, S2S, ends 2027-01-20 |
$32.00 per 1M tokens | |
gpt-realtime-1.5GPT-Realtime-1.5OpenAI, S2S |
$32.00 per 1M tokens | |
gpt-realtime-2GPT-Realtime-2OpenAI, S2S |
$32.00 per 1M tokens | |
gpt-realtime-2.1GPT-Realtime-2.1OpenAI, S2S |
$32.00 per 1M tokens | |
gpt-realtime-2.1-miniGPT-Realtime-2.1 miniOpenAI, S2S |
$10.00 per 1M tokens | |
gpt-realtime-2025-08-28GPT-Realtime (2025-08-28)OpenAI, S2S, ends 2027-01-20 |
$32.00 per 1M tokens | |
gpt-realtime-miniGPT-Realtime miniOpenAI, S2S, ends 2027-01-20 |
$10.00 per 1M tokens | |
gpt-realtime-mini-2025-12-15GPT-Realtime mini (2025-12-15)OpenAI, S2S, ends 2027-01-20 |
$10.00 per 1M tokens | |
stepaudio-2.5-realtimeStepAudio 2.5 RealtimeStepFun, S2S |
$1.50 per 1M tokens | |
stepaudio-3-realtime-previewStepAudio 3 Realtime (Preview)StepFun, S2S, beta |
– | |
1.2.1.1Doubao Realtime Voice O2.0 (Omni)Volcengine, S2S |
¥80.00 per 1M tokens | |
1.2.6.1Doubao Realtime Voice 3.0 (Seeduplex, full-duplex)Volcengine, S2S |
¥80.00 per 1M tokens | |
2.2.0.0Doubao Realtime Voice SC2.0 (Strong Character)Volcengine, S2S |
¥80.00 per 1M tokens | |
volc.service_type.10053Doubao Simultaneous Interpretation 2.0Volcengine, S2S |
¥80.00 per 1M tokens | |
glm-4-voiceGLM-4-VoiceZhipu AI (Z.ai), S2S |
¥80.00 per 1M tokens | |
glm-realtime-airGLM-Realtime-AirZhipu AI (Z.ai), S2S |
¥0.3 per minute | |
glm-realtime-flashGLM-Realtime-FlashZhipu AI (Z.ai), S2S |
¥0.18 per minute |
Speech-to-speech by feature
Speech-to-speech by language
- Afrikaans
- Arabic
- Azerbaijani
- Belarusian
- Bengali
- Bosnian
- Bulgarian
- Cantonese
- Catalan
- Chinese
- Croatian
- Czech
- Danish
- Dutch
- English
- Estonian
- Filipino
- Finnish
- French
- Galician
- German
- Greek
- Gujarati
- Hebrew
- Hindi
- Hungarian
- Icelandic
- Indonesian
- Italian
- Japanese
- Javanese
- Kannada
- Kazakh
- Korean
- Latvian
- Macedonian
- Malay
- Malayalam
- Marathi
- Norwegian
- Persian
- Polish
- Portuguese
- Punjabi
- Romanian
- Russian
- Slovak
- Slovenian
- Spanish
- Swahili
- Swedish
- Tagalog
- Tamil
- Thai
- Turkish
- Ukrainian
- Urdu
- Vietnamese
- ast
- ceb
- ky
- nan
- tg
- wuu