Retell AI pricing 2026: Per-minute costs & plan guide

Kenneth Pangan
Written by

Kenneth Pangan

Katelin Teen
Reviewed by

Katelin Teen

Last edited October 5, 2026

Expert Verified
Illustration of a presenter walking a colleague through a three-column pricing plan comparison

AI voice agents are everywhere these days. Businesses are looking for ways to automate customer calls, and these tools promise to handle everything from scheduling appointments to answering questions, 24/7. In this conversation, Retell AI often comes up, especially for companies with strong engineering teams who want to build a voice solution from scratch.

But before you jump in, you have to talk about the cost. This guide is here to give you a clear, no-nonsense breakdown of the Retell AI pricing model. We'll get into its per-minute fees, the extra costs that can pop up, and help you figure out what you'd actually end up paying.

And while voice AI is impressive, we'll also touch on why starting with text-based support (like email and chat) might be a smarter first move for getting a faster, more predictable return.

What is Retell AI?

So, what is Retell AI? Think of it as a powerful set of building blocks for developers to create AI that can talk on the phone. It’s not really for your average support manager or business owner; it's built for engineering teams who need total control over every little detail of a call.

People typically use it for things like:

  • Outbound sales calls to qualify leads

  • Inbound phone lines for customer service

  • Automated appointment scheduling and reminders

The platform is known for its techy features, like low-latency streaming that makes conversations feel less robotic and interruption handling so the AI doesn't just talk over your customers. It also lets you plug in different Large Language Models (LLMs) or phone carriers. This makes it a super flexible tool, but you absolutely need some serious technical skills to make it work.

Infographic showing how Retell AI works, which is important for understanding the Retell AI pricing model.
Infographic showing how Retell AI works, which is important for understanding the Retell AI pricing model.

A detailed breakdown of the Retell AI pricing model

Retell AI uses a pay-as-you-go model, and new accounts start with $10 in free credits. On the surface, that sounds great and flexible, but it can make budgeting a real challenge. Your final bill isn't one simple number; it’s a mix of several different costs that can change from month to month.

The core components of your per-minute cost

Every single minute of every call is priced based on three separate services working together: the voice, the AI "brain," and the phone connection itself.

ComponentPricing (per minute)Key Details
Conversation Voice Engine$0.055 + $0.015 - $0.04A flat $0.055 Retell voice infrastructure fee, plus text-to-speech at $0.015 for most voice providers or $0.04 for ElevenLabs.
Large Language Model (LLM)$0.0016 - $0.32+This is the AI doing the actual thinking. The cost swings wildly with the LLM you choose. GPT 5 nano is $0.0016 per minute, while GPT 6 Astra is $0.32 standard and $0.64 on the fast tier. Speech-to-speech models run $0.07 to $0.38.
Telephony~$0.015This is the fee for using Retell's phone service (which runs on platforms like Twilio). You can bring your own carrier to avoid this fee, but you'll just be paying your own provider instead.

Other costs that can sneak up on you

On top of those core per-minute fees, a few other charges can find their way onto your bill, making the final cost even harder to pin down.

  • Knowledge Base Usage: If you need your AI to pull answers from your documents, that's an extra "$0.005/minute". You get your first 10 knowledge bases for free, but after that, it's "$8/month" for each one.

  • Batch & Branded Calls: Running big outbound call campaigns has its own costs. You'll pay "$0.005" for every number you dial in a batch and another "$0.10" for every call you make with a branded caller ID.

  • Phone Number Rentals: A basic phone number is "$2/month". But if you want a verified number to look more legit and avoid spam filters, that is "$10/month".

  • Concurrency: You can run 20 calls at the same time for free. If your call volume is higher than that, each extra line for a simultaneous call costs "$8/month".

The enterprise plan for bigger teams

For larger teams, Retell has an Enterprise Plan. It adds custom MSA/DPA terms, SSO, role-based access control, and 24/7 support with a dedicated portal. This is a "Contact Sales" deal, meaning the price isn't public. Retell also lets you self-sign a BAA or DPA at no extra fee, so compliance paperwork alone doesn't force an Enterprise deal. You’ll have to negotiate, which adds another layer of complexity to the whole process.

Retell AI pricing page headline "True pay as you go", as taken from Retell AI
Retell AI pricing page headline "True pay as you go", as taken from Retell AI

Key features and limitations

It's only fair to talk about what Retell AI does well, but it's just as important to understand where it might not be the right fit, especially if you're not a developer.

Where Retell AI gets it right: Control and flexibility

For a technical team, Retell AI is a dream playground. It provides a solid API and WebSocket support, giving developers precise control over conversations as they happen. The option to bring your own LLM or phone carrier is a big win for companies that have very specific technical needs.

Features like warm transfers (passing a call smoothly to a human) and branded caller ID are genuinely useful for building out advanced sales or support systems. If you have the engineers, you can pretty much build any voice experience you can think of.

The challenges for business users

The platform’s greatest strengths are also its biggest weaknesses for non-technical teams.

  • It leans on developers. Retell has a dashboard with agent templates, but custom functions, webhooks, and API integrations are code work. A support manager will usually want an engineer nearby for anything beyond a basic agent, which can slow down changes.

  • Visual flows and testing exist, but they bill like production. Retell has a node-based conversation flow editor and a Simulation tab where an AI plays the caller and grades each run. Testing is billed at production rates, with no free sandbox, so a big batch of test cases adds to the bill.

  • The costs are unpredictable. This is probably the biggest hurdle for most businesses. With so many moving parts in the pricing, it's almost impossible to know what your bill will be at the end of the month. A sudden spike in call volume or a decision to use a more powerful LLM could make your costs shoot up without warning. That’s a tough situation for any department trying to stick to a budget.

Retell does get expensive fast but their quality is solid.

A simpler alternative: Predictable pricing for text-based support

While voice AI gets a lot of hype, the truth is that most businesses can get a bigger, faster win by automating their text-based support first. Just think about all the repetitive questions your team answers every day via email, web chat, and Slack.

This is where a tool like eesel AI offers a different approach. It’s built to be simple, user-friendly, and have predictable pricing. Instead of a confusing mess of per-minute fees, eesel AI pricing is one fixed monthly price sized to volume, where 1 credit is a ticket or a chat. Plans start at $299/month for 500 credits, with a free plan of 100 credits.

Here’s a quick comparison of the two approaches:

FeatureRetell AIeesel AI
Pricing ModelModular, pay-as-you-go (per minute)Fixed monthly credit plans (1 credit per ticket or chat)
Cost PredictabilityLow (Varies with usage and features)High (Fixed monthly price, from $299/month for 500 credits)
Target UserDevelopers / EngineersSupport Managers / Business Users
Setup ProcessDashboard templates and flows, with code for custom toolsRadically self-serve, go live in minutes
Core Use CaseVoice Agent AutomationText-Based Support Automation (Help Desk, Chat)
SimulationSimulation tab with AI-simulated callers (billed at production rates)Built-in simulation on historical tickets
This screenshot shows eesel AI
This screenshot shows eesel AI

What really makes eesel AI different is that it’s radically self-serve. A support manager can sign up, connect their help desk like Zendesk or Freshdesk in a single click, and have an AI agent up and running in minutes. No coding required.

Plus, eesel AI has a powerful simulation mode. You can test your AI on thousands of your own past support tickets to see exactly how it will perform before you ever let it talk to a real customer. Retell's simulation runs on scenarios you write; eesel's runs over your own past tickets, which gives a direct picture of your resolution rate.

The Retell AI pricing model is complex and really only makes sense for a specific type of user: a developer-led team with a flexible budget building a custom voice AI project. It’s powerful, but that power comes at the cost of simplicity and predictability.

For most businesses just looking for a fast, reliable, and easy way to lighten their support load, a complicated voice platform probably isn't the answer. Focusing on the text-based channels where your customers already are, with a tool like eesel AI, is often the smarter, more efficient choice that delivers a clear return without all the headaches.

Ready to see how simple AI support automation can actually be? Start your free eesel AI trial today.

Frequently asked questions

How does Retell AI pricing generally work for businesses?

Retell AI pricing operates on a pay-as-you-go model, meaning your costs are based on your usage each month. It's a modular system where various services contribute to the overall per-minute cost of a call.

What are the main components that make up the per-minute Retell AI pricing?

The core Retell AI pricing per minute is composed of three main parts: the Conversation Voice Engine (Retell's infrastructure fee plus text-to-speech), the Large Language Model (LLM) for the AI's "brain," and Telephony fees for the phone connection. Each component has its own rate that factors into the total.

Are there any hidden or additional fees I should be aware of when considering Retell AI pricing?

Yes, beyond the core per-minute charges, additional fees can include Knowledge Base usage, costs for Batch & Branded Calls, Phone Number Rentals, and Concurrency fees if you exceed the free simultaneous call limit. These can add up and affect your final bill.

Does Retell AI pricing offer an enterprise plan for larger companies?

Retell AI does offer an Enterprise Plan with custom pricing. It adds items like custom MSA/DPA terms, SSO, role-based access control, and 24/7 support with a dedicated portal, but the exact Retell AI pricing terms are negotiated directly with their sales team.

Why is it difficult for businesses to predict their monthly Retell AI pricing?

Predicting monthly Retell AI pricing is challenging because it's based on variable factors like call volume, the specific LLM chosen, and the utilization of various add-on features. Any change in usage or feature selection can significantly alter the final cost, making budgeting difficult.

How does Retell AI pricing compare to solutions for text-based support like eesel AI in terms of predictability?

Retell AI pricing is highly variable and depends on per-minute usage and multiple add-ons, making it less predictable. In contrast, eesel AI uses one fixed monthly price sized to volume, where 1 credit covers a ticket or a chat, with plans from $299/month for 500 credits and a free plan of 100 credits.

Share this article

Kenneth Pangan

Article by

Kenneth Pangan

Writer and marketer for over ten years, Kenneth Pangan splits his time between history, politics, and art with plenty of interruptions from his dogs demanding attention.

Related Posts

All posts →
Hand-drawn illustration of a person talking to a voice AI agent while a colleague points at a stopwatch with a dollar coin
Guides

GPT Realtime Mini pricing 2026: All per-minute costs

OpenAI's GPT realtime mini pricing looks affordable, but token-based billing is confusing. We ran the tests to give you a practical cost-per-minute breakdown and show you the hidden costs of building a production-ready AI voice agent.

KiraKiraOct 8, 2025
A guide to Zoom pricing in 2026: Plans, add-ons, and AI
Guides

Zoom pricing 2026: Every plan and add-on side by side

Navigating Zoom's pricing tiers can be complex. This guide provides a clear overview of Zoom Workplace plans, add-ons, and the AI Companion (now branded ZoomMate), helping you understand the costs and find the best value for your team's needs in 2026.

Rama AdiRama AdiSep 29, 2025
Hand-drawn hero in Town's rust red: an AI assistant ticking items on a HubSpot record while two people work at a laptop and on a phone call
Guides

Town AI for HubSpot: what it can update, the routines, and the plan you need

How Town AI works with HubSpot: what the connection can read and write, the five routines that keep deals current, the approval rules, and why Pro is the real plan.

Rama AdiRama AdiOct 6, 2026
Hand-drawn hero in Town's rust red: a person at a laptop reviewing a drafted follow-up email after a client call
Guides

Town AI for client follow-ups: routines, setup, and the plan you need

How to use Town AI for client follow-ups: the routines that draft recaps after calls and nudge quiet deals, how to set one up, and why Pro is the real starting plan.

Rama AdiRama AdiOct 6, 2026
Illustration of a team reviewing GPU cloud pricing charts next to a cloud server
Guides

CoreWeave pricing 2026: GPU costs per hour explained

Thinking about building on CoreWeave? Here's a detailed look at CoreWeave pricing for their AI cloud, plus a guide to the other critical AI costs you need to consider.

Kenneth PanganKenneth PanganOct 1, 2025
A person holding a blank price tag while another lifts a cloth off a hidden usage meter behind it
Guides

EasyAgentForYou pricing 2026: every plan, and the meter it doesn't show you

EasyAgentForYou costs $10, $25 or $70 a month. Those numbers buy agent slots and storage, not usage. The meter that sets your real bill is documented in the API reference at no published rate.

KiraKiraAug 19, 2026
Illustrated hero banner for a breakdown of Ayudo's pricing, showing per-ticket usage rates alongside per-seat plan cards
Guides

Ayudo pricing 2026: what $0.40 per ticket really costs

Ayudo pricing is $0.40 per ticket plus $0.05 per voice minute, with seats at $99 or $119. Here is what the page never defines, and what that costs you.

Kurnia KharismaKurnia KharismaAug 20, 2026
Hand-drawn illustration of a team reviewing an Asana task board beside the Asana logo
Guides

Asana pricing 2026: Every team plan compared in detail

Is Asana pricing right for your team? This guide breaks down every plan, from the free tier to Enterprise. Discover key features, user limits, and hidden costs like minimum seat requirements.

Rama AdiRama AdiSep 15, 2025
Wix e-commerce pricing 2026: A guide to plans & hidden costs
Guides

Wix ecommerce pricing 2026: Which plan to pick (and avoid)

Wondering about the true cost of a Wix online store? This guide breaks down the 2026 Wix e-commerce pricing plans, from the Light tier to Business Elite. We'll also uncover the hidden costs, like domain renewals, app fees, and the often-overlooked expense of scaling your customer support. Discover how to manage your total investment and maximize your store's profitability.

Kenneth PanganKenneth PanganSep 22, 2025

Ready to hire your AI teammate?

Set up in minutes. No credit card required.

Get started free