Products

AI Voice Agent Platform

Build LLM-powered AI Voice Agents

Explore more →

MRCP ASR

ASR for VoiceXML IVR Platforms

Explore more →

Developers

APIs, SDKs, Docs & Pricing

Explore more →

By Industry

Healthcare Payers

Help

Documentation

↗

Implementation guides

Status Page

↗

System health & uptime

Support

↗

Get help from the team

Build with Voicegain.

Speech-to-Text, Telephony Bot, Speech Analytics, and MRCP ASR. Cloud, your VPC, or on-prem. Sign up free with $50 of credit — no credit card, no demo call.

Talk to UsRead the docs →
What you can build

Four APIs. One platform.

Every Voicegain API shares the same auth, the same audio formats, and the same deployment options. Use one, or compose them.

Speech-to-text

Batch · Streaming

Speech-to-text API

89% accuracy on call audio. Omega for batch, Kappa for sub-100ms streaming. Diarization, PII redaction, custom vocabulary.

View docs →

Telephony Bot

SIP · Callbacks

Telephony Bot API

Build voice bots and IVRs by inviting Voicegain into a SIP session. JSON-driven turns. Bring your own LLM or business logic.

View docs →

Analytics

Async

Speech Analytics API

Sentiment, named-entity recognition, keywords, intent, and topic extraction. Transcription plus analysis in a single call.

View docs →

MRCP ASR

MRCPv1 · MRCPv2

MRCP ASR

Drop-in speech recognition for VoiceXML IVR. SRGS, grXML, JSGF grammars. Replaces legacy ASRs with no application rewrites.

View docs →
Get started

From signup to first transcript in three steps.

No sales call. Sign up, grab your API key, send a request.

01

Create a free account.

Email and password, or sign in with Google. No credit card required. You get $50 of platform credit on signup.

02

Grab a key from the console.

Project-scoped keys with their own rate limits and audit logs. Rotate or revoke any key without downtime.

03

Make your first call.

One request from your terminal — audio in, transcript out. Or open a streaming WebSocket for real-time recognition.

Code examples

Three common use cases. Python is our official SDK

*for other languages, call the REST and WebSocket API directly.

Streaming STT
Batch transcribe
Telephony Bot
from voicegain_speech import Configuration, ApiClient, AsrApi

cfg = Configuration()
cfg.access_token = os.environ["VG_KEY"]
api = AsrApi(ApiClient(cfg))

session = api.asr_recognize_async_post(
  async_recognition_request={
    "audio": {"source": {"stream": {"protocol": "WEBSOCKET"}}},
    "settings": {"asr": {"languages": ["en-US"], "model": "kappa"}}
  }
)
Models

Three models, trained on real call audio.

Voicegain owns its models — not wrappers on someone else's API. Pick the one that fits your use case.

Batch · best accuracy

Omega

Highest accuracy for batch and post-call processing. 89% on the Voicegain broad benchmark. 6 languages.

Streaming · low latency

Kappa

Real-time recognition under 100ms end-of-utterance. ~84% accuracy. The model Voicegain uses for live voice agents.

Multilingual

Voicegain-Whisper

Hosted Whisper-Large-V3 plus Voicegain fine-tunes. 15+ languages. ~86% accuracy. Best for non-English audio.

←→

Languages: English, Spanish, German, Portuguese, Korean, Hindi (+ Dutch & French in pipeline). Custom acoustic and language model training available.

Deployment

Run it where you need it.

Same APIs, same SDK, same models — in our cloud or your environment.

Voicegain Cloud

HIPAA-compliant, SOC 2-certified multi-tenant SaaS. Where Casey runs for mid-market health plans and TPAs, and the fastest start for Platform builders.

QUICK TO START

Your private cloud

Deploy in your AWS, Azure, or GCP VPC. Same APIs, same UIs — your network, your encryption keys.

FOR ENTERPRISE HEALTHCARE PAYERS

On-premises

Bare-metal Kubernetes or fully air-gapped. Your datacenter, your hardware, no internet egress. Perfect for regulated industries.

For Regulated Industries
Start free • $50 creditRead the docs →
Pricing

Free to start. Pay only for what you use.

No seat fees, no data-egress traps. Cloud is billed per second of audio; Edge deployments are licensed per port.

Free to start

$50 Credit

Enough to evaluate, build, and ship a real integration — no credit card required.

  • All Cloud APIs included
  • Full SDK & console access
  • In-house Voicegain SIP stack
Start free • $50 credit

Cloud

$0.00020 / sec

Per second of audio, 6-second minimum per request. Volume and term commits lower the rate.

  • Batch & streaming STT
  • Speech Analytics & Bot APIs
  • No infrastructure to manage
Start free • $50 credit

Edge

Per port / month

Licensed per port for deployments in your VPC or datacenter. Call-home or air-gapped.

  • Runs in your environment
  • Per-port throughput licensing
  • Air-gapped option available
Get a quote

Enterprise

Custom

For high volume, custom SLAs, dedicated support, custom model training, and BAAs.

  • Custom rate cards
  • Dedicated success engineering
  • SLA & uptime guarantees
Get a quote
FAQ

Common questions

Have more questions?

Talk to us →

How do I authenticate API requests?

+

Every request uses a bearer token generated from your project-scoped API key in the console — no OAuth dance required.

What audio formats are supported?

+

WAV, MP3, FLAC, and raw PCM are all supported for batch and streaming requests, at common telephony and broadcast sample rates.

Is there a free tier?

+

Yes — every account starts with $50 of platform credit, no credit card required, enough to build and ship a real integration.

Can I run Voicegain on-premises or air-gapped?

+

Yes — Edge licensing runs the same APIs and models in your VPC or datacenter, call-home or fully air-gapped.

Where to next

Keep building.

The full documentation lives on the Voicegain Support Center. The SDK and sample code live on GitHub. Here's everything in one place.

Documentation

Documentation

API reference, guides, and tutorials across STT, Telephony Bot, Analytics, MRCP, and Edge deployment.

support.voicegain.ai →

Open source

GitHub

The official Python SDK, the platform repo with dozens of sample integrations, and a transcript-comparison utility.

github.com/voicegain →
Trust & compliance

The certifications your procurement team will want.

HIPAA

Full compliance with healthcare regulations.

SOC 2 Type 2

Audit complete on all 5 criteria of SOC 2.

PCI DSS

Complies with Level-2 v4.0 PCI guidelines.

E2E-Encryption

Secure solutions from call start to finish.

Voice AI under
Your Control

Talk to usTry Voicegain →