Email Auto-Responder
Draft personalized reply emails in seconds
Automatically generate context-aware draft replies to incoming customer emails, pulling from your product knowledge base and the customer's history. Agents review and send, or the system sends directly for well-defined categories like order confirmations.
GPT-6 Luna
for 500K tokens a month · 65% input / 35% output
Why this model
For this workload — 500K tokens a month, 65% of them input — it comes to about $0.12 a month, the lowest of the models suggested here.
Alternative models
Implementation tips
- 1
Include 3–5 few-shot examples in your system prompt showing the tone and format expected — this alone cuts revision time by half compared to zero-shot.
- 2
Classify incoming emails into intent buckets (complaint, inquiry, refund) before drafting, so you can apply specialized sub-prompts per category.
- 3
Add a mandatory 'key facts' extraction step before writing the reply: the model lists the email's main points, ensuring nothing is missed in the response.