Anatomy of a voice agent call: $0.39 for 4 minutes, half platform fee, only 10% LLM
Warm-Reaction-456 · reddit · 2026-10-12
An automation agency itemizes a real 4-minute outbound voice agent call: Vapi $0.200, ElevenLabs Flash $0.075, Twilio $0.056, GPT-5.4 mini $0.040 (50k input/500 output tokens), Deepgram Nova-3 $0.019—about $0.39 total. Key takeaways: the platform fee is half the bill while the LLM is only 10%; input tokens grow faster than call length since every turn resends history; tool results (e.g. a 1,500-token CRM record) ride along all remaining turns—pass only needed fields; prompt caching cuts the LLM line from 4¢ to under 1¢ (and GPT-5.5 from 27¢ to 5¢); shorter agent replies cut per-character voice costs. Optimize for cost per real conversation, not per call.
More from coding & agent
- Claude cut a 2.6-hour video export to 20 minutes — but only after being told to rebuild the server — byhewar · 2026-10-23
- 3 flight-booking rules my AI agent checks that I'd never do by hand — Efistoffeles · 2026-10-23
- OpenJet claims Qwen3.8-27B coding runs 1.9x faster than llama.cpp on one RTX 3090 — JeremyCMorgan · 2026-10-23
- Building an autonomous SEO agent in Hermes with three skills: programmatic pages, Keywords Everywhere, DataForSEO — EXM7777 · 2026-10-23
- Building a sports betting AI agent with onchain execution and human approval by default: architecture and lessons — Dry-Comfortable-2678 · 2026-10-23
- How do you audit AI agent decisions in regulated workflows? A field survey — Ok_Way_1993 · 2026-10-23