Retell AI
Platform to build, test, deploy, and monitor AI voice agents for calls
Last reviewed 2026-06-18
Retell AI is a platform for building, testing, deploying, and monitoring AI voice agents that handle phone calls in production. A deployed agent fields inbound calls or runs outbound campaigns, qualifies leads, answers questions, and takes action in real time over telephony, conducting the live call end to end without a human on the line. The platform handles the real-time voice infrastructure (low-latency streaming, interruption and barge-in handling, turn-taking) and a voice-streaming API to connect agents to phone or web over WebSocket audio. To build agents, Retell offers Conversation Flow Agents (node-and-transition flows for structured, controllable calls) plus single- and multi-prompt agents for looser use cases. It is provider-flexible: bring your own telephony (Twilio, Telnyx, Vonage) or use Retell's built-in carrier, choose premium voices (ElevenLabs, PlayHT), and connect your own LLM. Designing and testing flows is human-driven; the running call is autonomous.
What it can do
Run live phone calls end to end
AutonomousA deployed voice agent conducts a full inbound or outbound phone conversation with low-latency streaming, interruption and barge-in handling, and turn-taking, without a human on the call.
sourceTake actions and call functions mid-call
AutonomousAgents call functions and connected tools during a conversation to look up data, qualify leads, and trigger actions in real time.
sourceBuild Conversation Flow Agents
SupervisedDesign structured calls as nodes and transitions for fine-grained control, or use single- and multi-prompt agents for looser use cases; flow design and tuning are human-driven.
sourceTest and monitor agents in production
SupervisedAutomated testing, call monitoring, and analytics for deployed agents; setup, review, and tuning are human-driven.
source
Strengths
- +Provider-flexible: bring your own telephony, voices, and LLM, or use Retell's built-in carrier
- +Structured Conversation Flow Agents (nodes and transitions) give fine control alongside looser prompt-based agents
- +Usage-based pricing with no mandatory base subscription and free credits to start; built-in testing and monitoring
Limitations
- −Advertised $0.07/min covers only the voice engine; LLM, telephony, and international add to real per-minute cost
- −Building reliable structured flows for complex calls takes meaningful human setup
- −Enterprise tier carries a high entry price relative to the self-serve usage rate
Overview
Retell AI is a platform for building, testing, deploying, and monitoring AI voice agents that handle phone calls in production, for both inbound and outbound use.
What it does
A deployed Retell agent runs a full live phone call: low-latency streaming, interruption and barge-in handling, turn-taking, and function calls to take action in real time. The live call runs autonomously. To build agents, Retell offers Conversation Flow Agents (nodes and transitions for structured control) plus single- and multi-prompt agents for looser use cases, a voice-streaming API over WebSocket, and built-in testing and monitoring. Flow design and tuning are human-driven.
Integrations & setup
Bring-your-own telephony (Twilio, Telnyx, Vonage) or Retell's built-in carrier; premium voices via ElevenLabs and PlayHT; connect your own LLM. Function calling lets agents reach connected tools mid-call.
Pricing
Usage-based with no mandatory base subscription and free starter credits. The advertised voice-engine rate is around $0.07/min, with the LLM and telephony billed separately on top, so real cost runs higher. A managed Enterprise plan reportedly starts around $8,000.
Traction
Retell AI is a Y Combinator (W24) company that raised a reported ~$4.6M seed round in 2024 led by Alt Capital with Y Combinator participating. Revenue and ARR figures circulating publicly are company-reported or third-party estimates, not audited.
Best for / not for
Best for developers and teams that want a flexible, pay-as-you-go voice agent platform with a choice of telephony, voices, and LLM. Less suited to teams wanting a single all-in flat per-minute price with no add-on math.
Alternatives
Bland AI runs an owned in-house voice stack; Vapi is a closely comparable API-first, model-agnostic voice platform.
What people are saying
We aggregate real LinkedIn discussion into sentiment for the agents people search most. Retell AI isn't tracked yet, want it added? Request tracking.
FAQ
How does Retell AI charge?+
Usage-based with no mandatory base subscription. The advertised voice-engine rate is about $0.07/min, with your chosen LLM and telephony billed on top, so realistic production cost is higher. A managed Enterprise plan reportedly starts around $8,000.
Is a Retell agent autonomous?+
On a live call, yes: a deployed agent runs the full phone conversation end to end, including function calls and actions, without a human on the line. Building and testing the conversation flows is human-driven and supervised.
Sources
- Retell AI (official site) · accessed 2026-06-18
- Retell AI pricing (official) · accessed 2026-06-18
- Retell AI seed funding announcement · accessed 2026-06-18
- Retell AI raises $4.6M to build AI agents (Y Combinator) · accessed 2026-06-18
Last reviewed 2026-06-18