Pricing Overview
SmarterAvatar uses transparent, usage-based pricing. You pay for the AI services you consume—no hidden platform fees.
Architecture Options
Choose the architecture that fits your infrastructure and requirements:
Option A: Full SmarterAvatar Stack (Recommended)
Use our LLM + RAG + Frontend
Best for: Quick deployment, no existing AI infrastructure
┌─────────────────┐ ┌─────────────────┐ ┌─────────────────┐
│ Your Docs │────▶│ SmarterAvatar │────▶│ HeyGen │
│ (S3/Upload) │ │ (Gemini + RAG) │ │ (Avatar) │
└─────────────────┘ └─────────────────┘ └─────────────────┘
Components:
- LLM: Google Gemini (with File Search RAG)
- STT: Whisper or Deepgram
- Avatar: HeyGen LiveAvatar
- Frontend: SmarterAvatar React App
- Analytics: Supabase
Per 3-minute session: ~$0.35
Option B: Hybrid - Your Backend + Our Frontend
Use your LLM/RAG (e.g., AWS Bedrock) + Our frontend
Best for: Existing AI infrastructure, data sovereignty requirements
┌─────────────────┐ ┌─────────────────┐ ┌─────────────────┐
│ Your Docs │────▶│ Your Backend │────▶│ SmarterAvatar │
│ (S3) │ │ (AWS Bedrock) │ │ (Frontend) │
└─────────────────┘ └─────────────────┘ └─────────────────┘
│
▼
┌─────────────────┐
│ HeyGen │
│ (Avatar) │
└─────────────────┘
Components:
- LLM: Your AWS Bedrock / FortiASK (you manage)
- RAG: Your Knowledge Base (you manage)
- STT: Whisper or Deepgram
- Avatar: HeyGen LiveAvatar
- Frontend: SmarterAvatar React App
- Analytics: Supabase
Per 3-minute session: ~$0.40 (varies by your LLM choice)
Cost Breakdown by Component
Per-Minute Costs
| Component | Provider | Cost/Minute |
|---|---|---|
| Avatar | HeyGen (Scale) | $0.10 |
| Avatar | HeyGen (Pro) | $0.198 |
| STT | Whisper | $0.006 |
| STT | Deepgram | $0.0077 |
| LLM | Gemini Flash | ~$0.003 |
| LLM | Claude Sonnet | ~$0.027 |
| LLM | AWS Bedrock (Claude) | ~$0.027 |
Per-Session Costs (3 minutes average)
| Configuration | Avatar | STT | LLM | Total |
|---|---|---|---|---|
| Budget (Gemini + Whisper) | $0.30 | $0.02 | $0.01 | $0.33 |
| Standard (Gemini + Deepgram) | $0.30 | $0.02 | $0.01 | $0.33 |
| Premium (Bedrock + Deepgram) | $0.30 | $0.02 | $0.08 | $0.40 |
Monthly Projections
Based on 3-minute average sessions:
| Sessions/Month | Budget Config | Premium Config |
|---|---|---|
| 1,000 | $330 | $400 |
| 5,000 | $1,650 | $2,000 |
| 10,000 | $3,300 | $4,000 |
| 25,000 | $8,250 | $10,000 |
Add infrastructure costs: ~$45-70/month (Supabase, hosting)
Infrastructure Costs
Fixed monthly costs regardless of session volume:
| Service | Cost | Purpose |
|---|---|---|
| Supabase Pro | $25/mo | Database, analytics |
| Resend | $0-20/mo | Email transcripts |
| Vercel | $0-20/mo | Hosting |
| Total | $25-65/mo |
BYOK Model
With BYOK (Bring Your Own Key), you pay providers directly:
You pay:
- HeyGen subscription/credits
- OpenAI/Deepgram usage
- Supabase subscription
We provide:
- Platform license
- Setup & integration
- Admin dashboard
- Ongoing support (optional)
This ensures you have full control and visibility into your AI spend.