WaaV Gateway - Supported Providers

Last Updated: 2026-01-17 Total Providers: 70+ Cloud API Providers Implemented: 46 | Blocked: 10 | Planned: 14

WaaV Gateway provides unified access to 70+ cloud-based Speech-to-Text (STT), Text-to-Speech (TTS), and Audio-to-Audio providers worldwide.

Quick Navigation

STT Providers (27 implemented)
TTS Providers (32 implemented)
Realtime/Audio-to-Audio Providers (2 implemented)
Provider by Region
Feature Comparison
Configuration Guide
Blocked Providers

Speech-to-Text (STT) Providers

Global Cloud Leaders

Provider	Protocol	Languages	Key Features	Env Variable
Deepgram	WebSocket	36+	Nova-2 model, real-time streaming, keyword boosting	`DEEPGRAM_API_KEY`
Google Cloud	gRPC	125+	Chirp 3, speaker diarization, word timestamps	`GOOGLE_APPLICATION_CREDENTIALS`
Microsoft Azure	WebSocket	100+	Custom speech models, pronunciation assessment	`AZURE_SPEECH_SUBSCRIPTION_KEY`
OpenAI	REST	57+	Whisper model, translation support, gpt-4o-transcribe	`OPENAI_API_KEY`
ElevenLabs	WebSocket	29+	Voice activity detection, multilingual	`ELEVENLABS_API_KEY`
AssemblyAI	WebSocket	99	Streaming API v3, end-of-turn detection, LeMUR AI	`ASSEMBLYAI_API_KEY`
Cartesia	WebSocket	15+	Ink-Whisper, ultra-low latency	`CARTESIA_API_KEY`
Amazon Transcribe	AWS SDK	100+	PII redaction, speaker diarization, streaming	`AWS_ACCESS_KEY_ID`
IBM Watson	WebSocket	30+	Smart formatting, custom acoustic models	`IBM_WATSON_API_KEY`
Groq	REST	57+	216x real-time speed, whisper-large-v3-turbo	`GROQ_API_KEY`

European Providers

Provider	Protocol	Languages	Key Features	Env Variable
Speechmatics	WebSocket	55+	Enhanced operating point, medical model	`SPEECHMATICS_API_KEY`
Gladia	WebSocket	100+	<300ms latency, code-switching, solaria-1	`GLADIA_API_KEY`
Rev AI	WebSocket	9 streaming	Speaker detection, custom vocabulary	`REVAI_API_KEY`
Phonexia	WebSocket	57-64	On-premises, voice biometrics	`PHONEXIA_TOKEN`
Acapela Group	REST	30+	Email/password auth, custom dictionaries	`ACAPELA_EMAIL`, `ACAPELA_PASSWORD`
Cereproc	REST	10+	Celtic languages, emotional voices	`CEREPROC_EMAIL`, `CEREPROC_PASSWORD`

Russia/CIS Providers

Provider	Protocol	Languages	Key Features	Env Variable
Yandex SpeechKit	REST	14+	Russian optimization, emotion detection	`YANDEX_API_KEY`
Tinkoff VoiceKit	gRPC	2	JWT auth, custom VAD, Russian/English	`TINKOFF_API_KEY`
SberDevices	REST	2	OAuth 2.0, SaluteSpeech, Russian/English	`SBERDEVICES_CLIENT_ID`

India Regional Providers

Provider	Protocol	Languages	Key Features	Env Variable
Sarvam AI	WebSocket	11+	Saarika v2.5, 22 Indian languages	`SARVAM_API_KEY`
Gnani.ai	gRPC/REST	14+	Voice biometrics, Indic languages	`GNANI_TOKEN`
Reverie	WebSocket	22+	Code-mixed Hinglish, hot-word boosting	`REVERIE_API_KEY`
AI4Bharat/Bhashini	REST	22+	Government ULCA API, free service	`BHASHINI_USER_ID`

China & East Asia Providers

Provider	Protocol	Languages	Key Features	Env Variable
iFlytek	WebSocket	18+	HMAC-SHA256 auth, Chinese dialects	`IFLYTEK_APPID`
Alibaba Cloud	WebSocket	25+	Qwen3-ASR, Paraformer, DashScope	`DASHSCOPE_API_KEY`
Baidu AI	WebSocket/REST	10+	Chinese dialects, emotion detection	`BAIDU_API_KEY`
Tencent Cloud	WebSocket	17	TC3-HMAC-SHA256, hotword boosting	`TENCENT_SECRET_ID`
Huawei Cloud	WebSocket	12	IAM token auth, Sichuan/Cantonese	`HUAWEI_PROJECT_ID`
NAVER CLOVA	REST	5	Korean optimization, NeuVis technology	`NAVER_CLIENT_ID`
AmiVoice	WebSocket	4	Japanese specialty, medical/finance	`AMIVOICE_APPKEY`

Southeast Asia Providers

Provider	Protocol	Languages	Key Features	Env Variable
Zalo AI	REST	1	Vietnamese, Northern/Southern accents	`ZALO_API_KEY`
FPT.AI	REST	1	Vietnamese, 7 voices	`FPT_API_KEY`
Viettel AI	REST	1	Vietnamese, 96% accuracy	`VIETTEL_TOKEN`
Prosa.ai	WebSocket/REST	1	Indonesian, stt-general models	`PROSA_API_KEY`
NECTEC	REST	1	Thai government AI, Partii engines	`NECTEC_API_KEY`

Text-to-Speech (TTS) Providers

Global Cloud Leaders

Provider	Protocol	Voices	Languages	Key Features	Env Variable
Deepgram	REST	10+	10+	Aura voices, natural prosody	`DEEPGRAM_API_KEY`
Google Cloud	gRPC	380+	75+	WaveNet, Neural2, Studio	`GOOGLE_APPLICATION_CREDENTIALS`
Microsoft Azure	WebSocket	400+	140+	Neural voices, SSML, emotion	`AZURE_SPEECH_SUBSCRIPTION_KEY`
OpenAI	REST	10	57+	TTS-1/TTS-1-HD, gpt-4o-mini-tts	`OPENAI_API_KEY`
ElevenLabs	WebSocket	1000+	29+	Voice cloning, emotional expression	`ELEVENLABS_API_KEY`
Cartesia	WebSocket	50+	40+	Sonic 3, <100ms latency, cloning	`CARTESIA_API_KEY`
Amazon Polly	AWS SDK	60+	30+	Neural/Generative engines, SSML	`AWS_ACCESS_KEY_ID`
IBM Watson	HTTP	30+	15+	V3 neural, rate/pitch control	`IBM_WATSON_API_KEY`

Voice Cloning Specialists

Provider	Protocol	Languages	Latency	Key Features	Env Variable
Hume AI	HTTP/WS	11+	~200ms	Octave, emotion control, 48 emotions	`HUME_API_KEY`
LMNT	HTTP	22+	~150ms	Voice cloning, top_p/temperature	`LMNT_API_KEY`
Play.ht	HTTP	36+	~190ms	PlayDialog multi-turn, cloning	`PLAYHT_API_KEY`
Murf.ai	HTTP	35+	~130ms	Falcon/Gen2, 12 regional endpoints	`MURF_API_KEY`
WellSaid Labs	HTTP	20+	~500ms	Caruso AI Director, 200+ avatars	`WELLSAID_API_KEY`
Resemble AI	HTTP	149+	~350ms	Chatterbox, paralinguistic tags	`RESEMBLE_API_KEY`
Speechify	HTTP	50+	~200ms	Simba models, 1000+ voices	`SPEECHIFY_API_KEY`
Unreal Speech	HTTP	8	~300ms	Kokoro voices, cost-effective	`UNREALSPEECH_API_KEY`
Smallest.ai	HTTP/WS	16	~100ms	Lightning model, Indian languages	`SMALLEST_API_KEY`

European Providers

Provider	Protocol	Voices	Languages	Key Features	Env Variable
Speechmatics	HTTP	4	1	English only, <200ms TTFA	`SPEECHMATICS_API_KEY`
Acapela Group	REST	250+	30+	17 audio formats, viseme data	`ACAPELA_EMAIL`
Cereproc	REST	20+	10+	Celtic languages, emotions	`CEREPROC_EMAIL`

Russia/CIS Providers

Provider	Protocol	Voices	Languages	Key Features	Env Variable
Yandex SpeechKit	REST	29+	4	6 emotions, Russian optimization	`YANDEX_API_KEY`
Tinkoff VoiceKit	gRPC	2	1	Alyona/Dorofeev, SSML	`TINKOFF_API_KEY`
SberDevices	REST	7	2	SaluteSpeech, SSML	`SBERDEVICES_CLIENT_ID`

India Regional Providers

Provider	Protocol	Voices	Languages	Key Features	Env Variable
Sarvam AI	REST	20+	11+	Bulbul model, Indian languages	`SARVAM_API_KEY`
Gnani.ai	REST	24+	12+	Multi-speaker, Indic languages	`GNANI_TOKEN`
Reverie	REST	36+	22+	Male/female per language	`REVERIE_API_KEY`
AI4Bharat/Bhashini	REST	44+	22+	Government API, free	`BHASHINI_USER_ID`

China & East Asia Providers

Provider	Protocol	Voices	Languages	Key Features	Env Variable
iFlytek	WebSocket	9+	15+	Speed/volume/pitch control	`IFLYTEK_APPID`
Alibaba Cloud	WebSocket	70+	25+	CosyVoice, Qwen3-TTS	`DASHSCOPE_API_KEY`
Baidu AI	REST	40+	10+	Chinese dialects, emotion	`BAIDU_API_KEY`
Tencent Cloud	REST	70+	17	Premium/emotional voices	`TENCENT_SECRET_ID`
Huawei Cloud	REST/WS	13+	2	Premium voices, text splitting	`HUAWEI_PROJECT_ID`
NAVER CLOVA	REST	100+	5	NeuVis neural, emotion control	`NAVER_CLIENT_ID`

Southeast Asia Providers

Provider	Protocol	Voices	Languages	Key Features	Env Variable
Zalo AI	REST	4	1	Vietnamese, Northern/Southern	`ZALO_API_KEY`
FPT.AI	REST	7	1	Vietnamese, MP3/WAV	`FPT_API_KEY`
Viettel AI	REST	12	1	Vietnamese, speed control	`VIETTEL_TOKEN`
Prosa.ai	REST	9	1	Indonesian, pitch/tempo	`PROSA_API_KEY`
NECTEC	REST	2	1	Thai, VAJA9 engine, free	`NECTEC_API_KEY`

Realtime Audio-to-Audio Providers

Provider	Protocol	Key Features	Use Case	Env Variable
OpenAI Realtime	WebSocket	GPT-4o, full-duplex, function calling, VAD	Conversational AI	`OPENAI_API_KEY`
Hume AI EVI	WebSocket	EVI 3, 48 emotions, empathic responses	Emotional AI	`HUME_API_KEY`

OpenAI Realtime Features

Full-duplex audio streaming with GPT-4o
Real-time voice activity detection
Function calling during conversation
Multiple voices: alloy, ash, ballad, coral, echo, fable, nova, onyx, sage, shimmer
WebSocket endpoint: wss://api.openai.com/v1/realtime

Hume EVI Features

48 emotion dimensions measured in real-time
Empathic response generation
Context continuity across utterances
Voice activity detection
EVI 3 (recommended), EVI 2 (deprecated)
WebSocket endpoint: wss://api.hume.ai/v0/evi/chat

Providers by Region

North America

Deepgram, OpenAI, AssemblyAI, Amazon (Transcribe/Polly), IBM Watson, Groq, ElevenLabs, LMNT, Play.ht, Hume AI, WellSaid Labs, Resemble AI, Speechify, Unreal Speech

Europe

Google Cloud, Microsoft Azure, Speechmatics (UK), Gladia (France), Rev AI, Phonexia, Acapela Group, Cereproc (Scotland)

Russia/CIS

Yandex SpeechKit, Tinkoff VoiceKit, SberDevices SaluteSpeech

India

Sarvam AI, Gnani.ai, Reverie, AI4Bharat/Bhashini, Smallest.ai

China

iFlytek (科大讯飞), Alibaba Cloud DashScope, Baidu AI, Tencent Cloud, Huawei Cloud

East Asia

NAVER CLOVA (Korea), AmiVoice (Japan)

Southeast Asia

Zalo AI (Vietnam), FPT.AI (Vietnam), Viettel AI (Vietnam), Prosa.ai (Indonesia), NECTEC (Thailand)

Feature Comparison

Speed Comparison (STT)

Tier	Providers	Typical Latency
Ultra-Fast	Groq (216x RT), Deepgram	<100ms
Fast	Gladia, AssemblyAI, Cartesia	100-300ms
Standard	Google, Azure, OpenAI	300-500ms

Speed Comparison (TTS)

Tier	Providers	Time-to-First-Audio
Ultra-Low	Smallest.ai (~100ms), Cartesia (<100ms), LMNT (~150ms)	<200ms
Low	Play.ht (~190ms), Murf.ai (~130ms), Hume AI (~200ms)	150-250ms
Standard	ElevenLabs, Azure, Resemble AI (~350ms)	250-500ms

Voice Cloning Support

Provider	Clone Time	Min Audio	Max Files	Features
ElevenLabs	~1 min	30s	N/A	Instant + Professional
Play.ht	~1 min	30s	N/A	Instant
LMNT	~1 min	5s	20 (250MB)	Enhancement option
Resemble AI	~1 min	10s	N/A	Rapid + Professional
Hume AI	~1 min	15s	N/A	Voice design
Cartesia	~1 min	10s	N/A	Instant

Emotion Control

Provider	Method	Emotions Supported
Hume AI	Natural language description	48 dimensions
ElevenLabs	Audio tags in text	5+ basic emotions
Azure	SSML prosody	10+ emotions
Tencent Cloud	Voice selection	10 emotional voices
NAVER CLOVA	Parameter	3 levels

Configuration Guide

Environment Variables

# Global Leaders
export DEEPGRAM_API_KEY="your-key"
export GOOGLE_APPLICATION_CREDENTIALS="/path/to/credentials.json"
export AZURE_SPEECH_SUBSCRIPTION_KEY="your-key"
export AZURE_SPEECH_REGION="eastus"
export OPENAI_API_KEY="your-key"
export ELEVENLABS_API_KEY="your-key"
export ASSEMBLYAI_API_KEY="your-key"
export CARTESIA_API_KEY="your-key"
export AWS_ACCESS_KEY_ID="your-key"
export AWS_SECRET_ACCESS_KEY="your-secret"
export IBM_WATSON_API_KEY="your-key"
export IBM_WATSON_INSTANCE_ID="your-instance"
export GROQ_API_KEY="gsk_your-key"

# Voice Cloning Specialists
export HUME_API_KEY="your-key"
export LMNT_API_KEY="your-key"
export PLAYHT_API_KEY="your-key"
export PLAYHT_USER_ID="your-user-id"
export MURF_API_KEY="your-key"
export WELLSAID_API_KEY="your-key"
export RESEMBLE_API_KEY="your-key"
export SPEECHIFY_API_KEY="your-key"
export UNREALSPEECH_API_KEY="your-key"
export SMALLEST_API_KEY="your-key"

# European Providers
export SPEECHMATICS_API_KEY="your-key"
export GLADIA_API_KEY="your-key"
export REVAI_API_KEY="your-key"
export ACAPELA_EMAIL="email"
export ACAPELA_PASSWORD="password"
export CEREPROC_EMAIL="email"
export CEREPROC_PASSWORD="password"

# Russia/CIS
export YANDEX_API_KEY="folder_id:api_key"
export TINKOFF_API_KEY="api_key|secret_key"
export SBERDEVICES_CLIENT_ID="client_id:client_secret"

# India Regional
export SARVAM_API_KEY="your-key"
export GNANI_TOKEN="your-token"
export GNANI_ACCESS_KEY="your-key"
export REVERIE_API_KEY="your-key"
export BHASHINI_USER_ID="user_id|ulca_api_key"

# China & East Asia
export IFLYTEK_APPID="app_id|api_key|api_secret"
export DASHSCOPE_API_KEY="your-key"
export BAIDU_API_KEY="api_key|secret_key"
export TENCENT_SECRET_ID="secret_id"
export TENCENT_SECRET_KEY="secret_key"
export HUAWEI_PROJECT_ID="your-project-id"
export HUAWEI_IAM_USER="username"
export HUAWEI_IAM_PASSWORD="password"
export NAVER_CLIENT_ID="your-client-id"
export NAVER_CLIENT_SECRET="your-secret"

# Southeast Asia
export ZALO_API_KEY="your-key"
export FPT_API_KEY="your-key"
export VIETTEL_TOKEN="your-token"
export PROSA_API_KEY="your-key"
export NECTEC_API_KEY="your-key"

WebSocket Configuration Examples

// Deepgram STT + ElevenLabs TTS
{
  "type": "config",
  "config": {
    "stt_provider": "deepgram",
    "tts_provider": "elevenlabs",
    "deepgram_model": "nova-2",
    "elevenlabs_voice_id": "21m00Tcm4TlvDq8ikWAM"
  }
}

// Groq STT (ultra-fast) + Hume AI TTS
{
  "type": "config",
  "config": {
    "stt_provider": "groq",
    "tts_provider": "hume",
    "groq_model": "whisper-large-v3-turbo",
    "hume_voice_id": "your-voice-id"
  }
}

// India: Gnani STT + Sarvam TTS
{
  "type": "config",
  "config": {
    "stt_provider": "gnani",
    "tts_provider": "sarvam",
    "language": "hi-IN"
  }
}

// China: Alibaba Cloud STT + TTS
{
  "type": "config",
  "config": {
    "stt_provider": "alibaba_cloud",
    "tts_provider": "alibaba_cloud",
    "alibaba_stt_model": "qwen3-asr-flash-realtime",
    "alibaba_tts_model": "cosyvoice-v3-flash"
  }
}

Blocked Providers

The following providers cannot be integrated due to API limitations:

Provider	Reason	Alternative
Otter.ai	Enterprise/Beta API only	AssemblyAI, Deepgram
Verbit	Enterprise workflow required	Rev AI, Speechmatics
SpeechText.AI	No real-time streaming	Groq, Deepgram
Speechly	Acquired by Roblox (2023)	AssemblyAI, Deepgram
ReadSpeaker	Enterprise only, no public API	Azure, Google Cloud
Nuance	EOL May 2027	Azure Speech Services
Kakao	Original API terminated (2022)	NAVER CLOVA
NTT COTOHA	Service terminated (2024)	Google Cloud, Azure
Vbee	API docs inaccessible	Zalo AI, FPT.AI
Kata.ai	Enterprise platform only	Prosa.ai
CoRover	Chatbot platform, no standalone API	Bhashini

Planned Providers

The following providers are planned for future releases:

Middle East & Africa

NeuralSpace (115+ languages)
Sestek/Knovvu (Turkey)
Lahajati (108 Arabic dialects)
AzReco (Azerbaijan)
ISSAI (Kazakh/Turkic)
Intron Health (African medical)
Lelapa AI (South Africa)
Lesan AI (Ethiopia)

Latin America

Vozy

Model Hosting Platforms

DeepInfra
Replicate

Specialized/Regional

AlfaNum (Balkans)
Aseto AI (Greek)
Elhuyar (Basque)
ABAIR (Irish Gaelic)
SignAll (ASL)
Signapse (BSL)
Botnoi Voice (Thailand)

Provider Selection Guide

Use Case	Recommended STT	Recommended TTS
Ultra-fast transcription	Groq, Deepgram	Cartesia, Smallest.ai
Low latency streaming	Deepgram, Gladia	LMNT, Play.ht, Cartesia
High accuracy	AssemblyAI, Google	Azure Neural, Google WaveNet
Voice cloning	-	ElevenLabs, Play.ht, Resemble AI
Emotion control	-	Hume AI, Azure SSML
Multi-language (global)	Google (125+), AssemblyAI (99)	Azure (140+), Google (75+)
Indian languages	Sarvam AI, Gnani, Bhashini	Sarvam AI, Reverie, Bhashini
Chinese languages	Alibaba, iFlytek, Tencent	Alibaba CosyVoice, iFlytek
Vietnamese	Zalo AI, FPT.AI, Viettel AI	Zalo AI, FPT.AI, Viettel AI
Cost-effective	Groq ($0.04/hr), Deepgram	OpenAI, Deepgram
Enterprise/HIPAA	Azure, Google, AWS, IBM	Azure, Google, AWS, IBM
Conversational AI	-	OpenAI Realtime, Hume EVI
On-premises	Phonexia	-

Version History

2026-01-17: Added NECTEC (Thailand), updated provider counts
2026-01-14: Added Tencent Cloud, Huawei Cloud, NAVER CLOVA, AmiVoice
2026-01-13: Added 20+ providers including Southeast Asia region
2026-01-07: Added Hume AI, LMNT, Play.ht
2026-01-06: Initial release with 25 providers

For detailed API documentation and implementation guides, see the provider documentation directory.

Uh oh!

FilesExpand file tree

SUPPORTED_PROVIDERS.md

Latest commit

History

SUPPORTED_PROVIDERS.md

File metadata and controls

WaaV Gateway - Supported Providers

Quick Navigation

Speech-to-Text (STT) Providers

Global Cloud Leaders

European Providers

Russia/CIS Providers

India Regional Providers

China & East Asia Providers

Southeast Asia Providers

Text-to-Speech (TTS) Providers

Global Cloud Leaders

Voice Cloning Specialists

European Providers

Russia/CIS Providers

India Regional Providers

China & East Asia Providers

Southeast Asia Providers

Realtime Audio-to-Audio Providers

OpenAI Realtime Features

Hume EVI Features

Providers by Region

North America

Europe

Russia/CIS

India

China

East Asia

Southeast Asia

Feature Comparison

Speed Comparison (STT)

Speed Comparison (TTS)

Voice Cloning Support

Emotion Control

Configuration Guide

Environment Variables

WebSocket Configuration Examples

Blocked Providers

Planned Providers

Middle East & Africa

Latin America

Model Hosting Platforms

Specialized/Regional

Provider Selection Guide

Version History