Retell AI Review (2026)
Bundled per-minute pricing with a managed stack, positioned as the production-ready option against the assemble-it-yourself platforms.
Some links on this page can be affiliate links. We may earn a commission at no cost to you. This never affects our rankings. How we stop it.
What a minute costs
The model →- Platform $0.07
- Language model $0.0108
- Telephony $0.014
Computed, not measured. The advertised rate plus the published list price of every component it excludes. Component rates are dated in the cost model. It is a floor — list pricing, no committed-use discount, no retries. Vendor pricing page and its own cost calculator, re-verified 2026-08-06. The page now advertises $0.07-$0.31 a minute, and the $0.07 floor decomposes exactly as Retell Voice Infra $0.055 plus the cheapest platform voice $0.015, with no language model on it. The model is a separate published line and the calculator's own default state shows $0.04/min for it. Speech-to-text sits inside the infra line; telephony is $0.015 on Retell's own carrier..
Retell AI in Depth
Retell bundles its own voice infrastructure and speech recognition into one rate. It does not bundle the language model, and the pricing page has stopped implying otherwise: it now advertises $0.07 to $0.31 a minute rather than a single figure.
The $0.07 floor decomposes exactly, on the vendor's own component table: Retell Voice Infra at $0.055 plus the cheapest platform voice at $0.015. Nothing else. It is reachable only by bringing your own model and your own carrier and paying both of them outside Retell's invoice.
Drive the vendor's own calculator and it says so itself. At its untouched defaults it returns $0.11 a minute, itemised as $0.04 of model, $0.055 of infrastructure and $0.015 of voice. Set every menu to its most expensive option and it returns $0.27.
What bundling actually buys
Not the price. The time.
A bring-your-own-key platform quotes $0.05 and requires you to hold four vendor relationships, tune endpointing across four seams, and diagnose failures that occur between components rather than inside them. Retell quotes more and hands you a working agent.
For a team shipping this quarter, the engineering time saved dwarfs the per-minute difference. For a team at serious volume, the arithmetic inverts, and the question is where.
Where the crossover sits
Take Vapi at $0.05 advertised, which computes to $0.0965 all-in on published component rates. Retell at $0.07 computes to $0.0948. On list pricing they land a fifth of a cent apart, which is a more interesting answer than either headline suggests.
So the honest framing is not bundled versus assembled on price. It is whether you will do the component optimisation. Teams that will, save meaningfully at volume. Teams that intend to and never get round to it pay more than they would have with the bundle.
The trade you are making
You cannot swap a component when something better ships. When a cheaper synthesis model appears, or a faster inference provider, a bring-your-own-key stack absorbs it in an afternoon and a bundled one waits for the vendor.
In a category moving this fast that is not a small thing, and it is the strongest argument against bundling regardless of what the arithmetic says today.
Setting Retell AI Up
The fastest path to a working phone agent among the developer-facing platforms. Self-serve, free credit, and an agent taking calls the same day without holding four vendor accounts.
- Test interruption handling before anything else. Call your own agent and talk over it. This is where deployments fail and where platforms differ most, and it takes ten minutes to evaluate.
- Check the latency you actually experience, mouth to ear, not the figure in the documentation. The published number is time to first byte and the caller hears something else.
- Add telephony to your model. The advertised rate excludes it, so budget the carrier cost separately from the start.
What the AI actually does
2 of 4 can act without a person in the loop. Those are the ones to scope carefully.
Voice agentsHolds a phone conversation with your customer$0.07-$0.31/min advertisedActs
The product. Retell's default is its own speech-to-speech engine at $0.055 a minute, and everything else on the rate card stacks on top of it.
Model choiceDecides how much a minute costs$0.003 to $0.345/minDrafts
Retell publishes a per-minute price for every model it offers, which almost nobody does. Twenty-plus of them, from GPT 5 nano at $0.003 a minute to GPT 5.5 at $0.16, with a separate faster tier at roughly double each. Claude 4.6 Sonnet is $0.08; Gemini 2.5 Flash Lite is $0.006.
That is a fifty-fold spread on one dropdown, and it is the single biggest lever on your bill.
Chat agentsThe same thing in text$0.001-$0.052 per AI messageActs
Priced per message rather than per minute, and again per model: GPT 5 nano at a tenth of a cent, GPT 5.5 at 5.2 cents. SMS is a flat $0.01 a message on top.
Safety add-onsGuardrails, PII removal, quality assurance+$0.005 to +$0.10/minScores
Worth reading as a group, because these are the lines that make an unattended voice agent defensible and every one is an extra: Safety Guardrails +$0.005/min, PII Removal +$0.01/min, Advanced Denoising +$0.005/min, and AI Quality Assurance at $0.10 a minute after the first hundred free.
Quality assurance alone costs more than the advertised floor of the whole product.
Retell publishes the component rate card this site normally has to build itself. Voice infrastructure, every text-to-speech voice, every model, telephony and each add-on all carry their own per-minute price on the pricing page. For a site whose whole method is adding up what a headline rate excludes, that is worth saying plainly: Retell does the disclosure most of this category does not.
And one supported configuration costs more than the advertised ceiling. The page quotes $0.07-$0.31 a minute. It also lists GPT Realtime as an alternative speech-to-speech engine at $0.345 a minute — for that layer alone, before telephony or add-ons. A buyer who picks it is above the top of the advertised range on one line item.
Strengths and Weaknesses
- The smallest gap between advertised and all-in cost of any platform here, at 1.2x against 2.4x for the bring-your-own-key options.
- Genuinely fast to production, with one vendor relationship instead of four.
- Bundled stack removes the tuning-across-seams problem that consumes the first month of assembled deployments.
- Self-serve with free credit, so the evaluation costs nothing.
- You cannot swap components. When something cheaper or better ships, you wait for the vendor rather than adopting it.
- At high volume with optimised components, assembled wins, though only if you actually do the optimising.
- Telephony is still excluded from the advertised rate, as it is everywhere in this category.
- Less control over the voice and model than teams with a specific quality requirement will want.
Retell AI Pricing
Published, self-serve, pay as you go, and now quoted as a range: $0.07 to $0.31 a minute. Speech recognition is inside the infrastructure line. The model and the telephony are not, and there are seven per-minute add-ons beyond them.
Computed against this site's cost model, the all-in floor is $0.0948. Set against the alternatives on published rates:
| Platform | Advertised | Computed all-in | Multiple |
|---|---|---|---|
| Retell | $0.07 | $0.0948 | 1.35x |
| Vapi | $0.05 | $0.1217 | 2.4x |
| Millis | $0.02 | $0.0665 | 3.3x |
Read that table carefully, because it inverts the usual assumption. The platform with the highest advertised rate has the lowest computed cost, and the one advertising $0.02 computes to nearly the same figure as the one advertising $0.07. Headline rates in this category are close to uninformative.
The Verdict
The cheapest voice platform on published rates despite advertising the highest number, and the right default unless you will genuinely optimise your own component stack.
- You want a working phone agent this week rather than this quarter
- You would rather hold one vendor relationship than four
- Nobody on the team wants to tune endpointing across component seams
- You are early enough that engineering time costs more than per-minute rates
- You are at volume and will genuinely do component optimisation
- You need a specific voice or model the bundle does not offer
- Swapping in cheaper components as they ship matters to your economics
- You need on-premise or self-hosted deployment
Can it take a payment?
Full tracker →Books meetings or resolves the call and hands off. Taking money is out of scope for the product as sold.