Plans
Pay-As-You-Go
$0 to start — pay only for what you use.
- $10 free credit on sign up
- 5 concurrent calls included
- All LLM providers (OpenAI, Anthropic, Gemini, Groq, Mistral)
- 40+ voice models
- No-code agent builder & templates
- PBX/SIP integration
- Batch calling campaigns
- Call analytics & recordings
- Webhooks & REST API
- Bring your own API keys
- Cancel anytime
Enterprise
Custom pricing — built for MSPs, agencies, and high-volume operations.
- Unlimited concurrent calls
- Unlimited knowledge base files
- Volume discounts up to 40% off platform fee
- Priority support + dedicated success manager
- Service level agreement (SLA — 99.9% uptime)
- White-label & reseller options
- Custom hosting & deployment
- SSO & SAML integration
- Free migration from existing systems
- Custom engineering extensions
Platform Fee Schedule
The platform fee applies to all plans and covers VoiceInfra infrastructure, telephony, call routing, and real-time WebSocket management. AI model costs (LLM, STT, TTS) are separate and billed at provider rates.| Item | Rate |
|---|---|
| Voice AI | $0.05 / min |
| Chat AI | $0.005 / message |
| Extension PBX add-on | +$0.015 / min |
| WhatsApp calls add-on | +$0.035 / min |
| Call transfers | $0.025 each |
| WhatsApp to PBX transfer | $0.50 each |
| Batch calling | $0.008 / dial |
| Extra concurrency | $8 / channel / month |
| Extra storage | $8 / 10 MB / month |
Volume discounts of up to 40% apply to the platform fee for enterprise accounts with $2,000+ in monthly usage. AI model costs are always passed through at provider rates — no markup, regardless of volume.
Real Cost Breakdown
Your total per-minute cost depends on the LLM, STT, and TTS models you choose. Here are three real production configurations across our customer base.| Tier | Rate | Stack | Example (30-min call) |
|---|---|---|---|
| Budget | $0.07 / min | GPT-4.1 Nano + Scribe V1 (ElevenLabs) + Deepgram TTS | ~$2.10 |
| Average | $0.10 / min | GPT-4.1 Mini + Nova 3 (Deepgram) + Eleven Flash V2.5 | ~$3.00 |
| Premium | $0.14 / min | GPT-4o + Nova 3 (Deepgram) + ElevenLabs Multilingual V2 | ~$4.20 |
Where Your Dollar Goes
For a typical 1-minute call at the $0.10/min average tier, costs break down like this:| Component | Cost | Share | What It Covers |
|---|---|---|---|
| Platform fee | $0.050 | 50% | Infrastructure, telephony, routing, WebSocket management |
| TTS (Text-to-Speech) | $0.030 | 30% | Converts AI responses to natural speech |
| LLM (AI brain) | $0.010 | 10% | Processes caller input and generates responses |
| STT (Speech-to-Text) | $0.010 | 10% | Transcribes caller speech in real time |
What’s Included in Every Plan
Both plans include the full VoiceInfra feature set. There is no feature gating between Pay-As-You-Go and Enterprise.All 5 LLM Providers
OpenAI, Anthropic Claude, Google Gemini, Groq, and Mistral. Assign different models to different agents and switch providers anytime — zero migration required.
40+ Voice Models
ElevenLabs, Cartesia, Rime Labs, and Deepgram voices with support for 30+ languages. Preview and compare voices directly in the agent builder.
No-Code Agent Builder
Guided wizard with industry-specific templates for customer support, appointment booking, sales qualification, and more. Auto-generates system prompts — no prompt engineering needed.
PBX & SIP Integration
Register AI agents as extensions on any SIP-compatible PBX (3CX, Yeastar, FreePBX, Asterisk, and more). Zero hardware changes, one IP to whitelist.
Call Analytics Dashboard
Real-time transcripts, sentiment analysis, call recordings, AI-generated summaries, and per-agent performance metrics for every call.
Developer API & Webhooks
REST API for dynamic outbound calls, real-time webhooks on call completion, custom functions, and MCP server integrations for any backend system.
Frequently Asked Questions
Do I need a credit card to start?
Do I need a credit card to start?
No. Sign up at voiceinfra.ai/signup and you receive 0.07/min). You only need a payment method after your free credit is exhausted.
Can I bring my own AI API keys?
Can I bring my own AI API keys?
Yes. Connect your own OpenAI, Anthropic, Google, Groq, or Mistral API keys in the Integrations section of your dashboard. When you use your own keys, AI model costs (LLM, STT, TTS) are charged directly by your provider — not through VoiceInfra. You still pay the VoiceInfra platform fee ($0.05/min), but you get full cost transparency and can leverage any existing API credits or volume agreements you have with providers.
Are there volume discounts?
Are there volume discounts?
Yes. Enterprise accounts with $2,000 or more in monthly platform usage qualify for volume discounts of up to 40% off the platform fee. AI model costs are always passed through at provider rates with no markup, regardless of volume. Contact the sales team at voiceinfra.ai/sales to discuss your expected volume and receive a custom quote.
What does the Enterprise plan include?
What does the Enterprise plan include?
Enterprise includes everything in Pay-As-You-Go plus: unlimited concurrent calls, unlimited knowledge base file storage, volume discounts up to 40% off the platform fee, a 99.9% uptime SLA, a dedicated success manager, direct access to the engineering team, white-label and reseller options for MSPs, custom hosting and deployment, SSO/SAML integration, and free migration from your existing systems. VoiceInfra’s team will also custom-build and scale AI agents to match your specific workflows. Reach out at voiceinfra.ai/sales to start the conversation.
How are WhatsApp calls priced?
How are WhatsApp calls priced?
WhatsApp voice calls use the same base platform fee (0.035/min**. All AI model costs (LLM, STT, TTS) remain identical to regular voice calls. A typical 1-minute WhatsApp call costs between 0.165/min depending on your model selection. If a caller transfers from WhatsApp to a PBX extension, an additional one-time fee of $0.50 applies per transfer.