Always the newest models — fetched live
The AI landscape moves fast, and yesterday's flagship model is tomorrow's budget option. That's why supportif.ai doesn't hardcode a model list: it queries the OpenAI and Anthropic APIs live and shows you every model currently available on your account. When a new model launches, it appears in your settings the same day — no app update required. You pick one model for triage and drafting, and you can change it whenever you like; the switch takes effect for the next incoming email.
Local LLM support
For merchants with strict data-locality requirements — or those who simply want to minimize costs — supportif.ai can talk to a locally hosted LLM via an OpenAI-compatible endpoint. Your customer emails never leave your infrastructure for inference, and on the Free plan, local LLM usage carries no markup at all: you pay nothing to us for it. This is a popular option for merchants running models on their own hardware or a private cloud.
Bring your own keys on Pro
On the Pro plan ($99/month) you enter your own OpenAI and/or Anthropic API keys. All AI calls then run directly against your accounts:
- No usage fees and no markup — you pay your provider directly at their raw rates.
- Your rate limits, your negotiated pricing — enterprise agreements with OpenAI or Anthropic apply automatically.
- Predictable costs — one flat $99/month to us, everything else on your provider invoice.
For high-volume shops this is almost always cheaper than the Free plan's cost-plus-30% model — our pricing page includes a break-even example.
Quality without lock-in
Different models have different strengths: some excel at nuanced, empathetic replies, others at fast and cheap classification. Because switching is instant and free, you can experiment: run a premium model for customer-facing drafts and observe how confidence scores and edit rates change. The app's learning system — style guide and solution patterns — is model-independent, so nothing you've taught supportif.ai is lost when you switch.