vsref

OpenAI Realtime API Review

Frontier speech-to-speech model API (gpt-realtime family) for low-latency voice agents over WebRTC, WebSocket, and SIP.

See pricing

Facts verified Jul 30, 2026Try OpenAI Realtime API →

If you sign up through links on this page, vsref may earn a commission; programs exist on both sides of most comparisons, and commissions never change verdicts. How we make money

What we know about OpenAI Realtime API

OpenAI Realtime API is a voice ai platform: frontier speech-to-speech model API (gpt-realtime family) for low-latency voice agents over WebRTC, WebSocket, and SIP. This profile tracks every OpenAI Realtime API fact we have verified, each linked to a primary source and dated.

OpenAI Realtime API does not publish a public per-minute rate we have been able to verify, so the price figure here stays blank until we can confirm one. We would rather show nothing than a guessed number; if you have a current OpenAI Realtime API quote, it is the fastest way for us to close that gap.

On capabilities, OpenAI Realtime API covers native sip trunking, interruption handling (barge-in), api-first (full lifecycle via api), and self-serve signup, and does not offer llm/tts costs passed through at cost?, annual contract required for best pricing?, and bring-your-own llm. Each of those is verified against OpenAI Realtime API's own docs or dashboard, not marketing copy.

In total we track 10 verified facts for OpenAI Realtime API today, each linking the primary source it came from so you can check our work - and vendor claims we have not measured ourselves are labeled as such on the OpenAI Realtime API fact sheet below.

Reviewed by vsref Editorialfacts verified Jul 30, 2026Methodology →

Fact sheet

Pricing
Pricing facts
LLM/TTS costs passed through at cost?✗ NoJul 30
Annual contract required for best pricing?✗ NoJul 30
Capabilities
Capabilities facts
Native SIP trunking✓ YesJul 30
Bring-your-own LLM✗ NoJul 30
Bring-your-own TTS voice✗ NoJul 30
Buy/port numbers in-platform✗ NoJul 30
Interruption handling (barge-in)✓ YesJul 30
Build experience
Build experience facts
API-first (full lifecycle via API)✓ YesJul 30
Native integrations count (+ key ones: GHL, HubSpot, Zapier, Make, Twilio)WebRTC, WebSocket, SIP; MCP + tool callingJul 30
Commercial
Commercial facts
Self-serve signup✓ YesJul 30

Considering a switch? Best OpenAI Realtime API alternatives →

TTS in this stack

A voice agent's voice quality and per-minute cost come from its text-to-speech engine. See how dedicated engines compare on price, latency, and cloning rights: ElevenLabs, Cartesia, OpenAI TTS, or the full text-to-speech comparison →

STT in this stack

A voice agent hears through its speech-to-text engine, and transcription accuracy and streaming latency are priced per audio minute. Compare the engines builders pair with OpenAI Realtime API: Deepgram, AssemblyAI, GPT-4o Transcribe, or the full speech-to-text comparison →

OpenAI Realtime API head-to-head

OpenAI Realtime API vs Vapi →won 0 · lost 4 · tied 1
Custom pipeline controllostEcommerce support and orderslostHealthcare compliancetieOutbound sales campaignslostSMB receptionist and bookinglost
OpenAI Realtime API vs ElevenLabs Agents →won 0 · lost 3 · tied 2
No-code agency deployslostCustom pipeline controllostHealthcare compliancetieOutbound sales campaignstieSMB receptionist and bookinglost
OpenAI Realtime API vs Google Gemini Live API →won 3 · lost 0 · tied 3
No-code agency deploystieCustom pipeline controlwonEcommerce support and orderswonHealthcare compliancetieOutbound sales campaignswonSMB receptionist and bookingtie
OpenAI Realtime API vs Hume EVI →won 0 · lost 4 · tied 2
No-code agency deployslostCustom pipeline controllostEcommerce support and orderstieHealthcare compliancelostOutbound sales campaignstieSMB receptionist and bookinglost