Seven voice AI pricing pages read directly

Vapi alternatives and Bland AI alternatives priced by the real all-in minute

Every Vapi alternative advertises a per-minute rate, and almost none of those rates are the bill. We read the pricing pages of Vapi, Retell AI, Bland AI, Telnyx, ElevenLabs, LiveKit and Synthflow, added up what each charges for the same phone agent, and found where the cheap headline gets expensive.

Vendor pricing pages read directly · No paid placements · Hire a voice AI developer from the form

Hire brief generator

demo · one brief per browser

describe the automation

hire

stack

That was the one free brief for this browser.

To hire for the job, put it on the public BotGigs job board: $99 a month on Growth, applications straight to your inbox, cancel any time.

Post a job, $99 a month

for a different task

Best-effort estimate, not a quote or a match

We could not scope your task just now, so this is an example brief for a different automation, shown so you can see the format. Your free brief has not been used: try again in a minute.

scope of work

who to hire

Screen for

Effort estimate

questions to ask your hire

Like the brief? Put the job in front of specialists on the board.

Post this job, $99 a month

the short answer

The strongest Vapi alternatives are Telnyx (about $0.056 a minute all in, by its own statement), LiveKit Cloud (roughly $0.06 to $0.10 once you add a model), Retell AI (about $0.13 on its cheapest listed model) and Bland AI ($0.14 a connected minute with the model bundled). Vapi itself charges $0.05 for hosting, but its own page lists transcription, model, voice and telephony on top, which puts a real agent at $0.09 to $0.14. So the headline rate predicts almost nothing: Vapi and Telnyx both advertise $0.05 and produce different bills. The costs that actually move a budget are fixed ones: Vapi HIPAA at $2,000 a month, Bland Build at $299, and Synthflow, whose pricing page now lists only annual contracts from $30,000.

The audit

What each Vapi alternative publishes on its own pricing page

Most Vapi alternatives articles are written by one of the alternatives, and each one ranks itself first. We are not a voice platform, so there is nothing here to rank. Every cell below was taken from the vendor pricing page itself. Rates change often, so check the page before you sign anything.

Platform	Published rate	What the rate leaves out	Who it suits
Vapi	$0.05 a minute hosting. Core plan $29 a month, Pro from $999 a month, HIPAA add-on $2,000 a month, extra concurrent lines $10 each	Transcription $0.0095 to $0.0099, model $0.0077 to $0.0452, voice $0.0146 to $0.0238, telephony $0.008 to $0.014 a minute. Pay as you go includes 4 concurrent calls	Developers who want to choose and swap every component
Retell AI	$0.055 a minute infrastructure, stated range $0.07 to $0.31. First 20 concurrent calls free, then $8 a line a month	Voice $0.015 ($0.040 for ElevenLabs voices), model $0.048 for Gemini 3.5 Flash (the cheapest listed now that GPT 4.1 is gone) up to $0.16 for GPT 5.5, US telephony $0.015, plus add-ons such as knowledge base and PII removal	Teams that want a managed stack with every line itemized
Bland AI	Start: no monthly fee, $0.14 a connected minute, 10 concurrent, 100 calls a day. Build: $299 a month, $0.12 a minute, 50 concurrent, 2,000 calls a day	Very little. Model, transcription and voices are included with no token charges. Transfers cost $0.05 (Start) or $0.04 (Build) a minute	Outbound or high call counts where one line item matters more than tuning
Telnyx	$0.05 a minute covering orchestration, transcription and voice. Vendor states about $0.056 all in for a production agent	Model tokens (about $0.004 a minute on an open model Telnyx hosts, pass-through for OpenAI or Anthropic) and telephony from $0.0032 a minute for US local inbound	Teams that want carrier, numbers and agent on one invoice
ElevenLabs Agents	Plans from $6 to $990 a month including 75 to 12,375 minutes, then $0.08 a minute, doubling to $0.16 when you exceed concurrency	The model and telephony are billed separately on top, at cost. BAAs for HIPAA only on Enterprise	Voice quality first, and agents that must speak many languages
LiveKit Cloud	Ship $50 a month with 5,000 agent session minutes, Scale $500 with 50,000, then $0.01 a minute	Transcription $0.0058, voice $0.030 and SIP telephony $0.004 a minute ($0.003 on Scale), plus the model. You write the agent in code	Engineering teams who want the lowest unit cost and full control
Synthflow	The pricing page lists one tier: Enterprise, from $30,000 a year. The older $29 to $449 monthly plans are no longer listed	No per-minute rate published. Price is scoped on volume, concurrency, telephony, integrations and security	Large deployments with a contract budget and a procurement process

Two columns tell the story. The published rate column makes Bland AI look the most expensive platform on the list at $0.14. The column next to it shows why that is misleading: Bland includes the model and the voice, while Vapi, Retell, Telnyx, ElevenLabs and LiveKit bill them separately. The only way to compare these platforms is to build the same agent on paper for each one, which is what the next section does. Synthflow has priced itself out of the comparison for most small buyers, and we cover what replaces it in Synthflow alternatives after the $30,000 minimum. Voiceflow is the other builder buyers ask about: it charges the same $0.05 platform minute as Vapi and bills model tokens at 1.25 times list, which puts a Haiku class phone agent near $0.11 to $0.13 all in, as worked through in Voiceflow pricing per message and per minute.

The all-in minute

One phone agent, priced on every platform

The agent: inbound US phone calls, a GPT 4.1 class model, a standard voice. Where a vendor bills the model or carrier at cost without printing a number, we used the model range Vapi publishes ($0.0077 to $0.0452) and its telephony range ($0.008 to $0.014) as the stand in, and say so. Monthly figures assume 5,000 minutes, roughly 1,700 three minute calls.

Platform	How the number is built	All-in per minute	At 5,000 minutes a month
Telnyx	$0.05 platform, model on top, US inbound carrier from $0.0032. Vendor figure with its hosted open model is $0.056	$0.056 (open model) to about $0.098 (OpenAI class)	About $280 to $490
LiveKit Cloud (Ship)	$0.01 session + $0.0058 transcription + $0.030 voice + $0.004 SIP + model. 5,000 session minutes are inside the $50 plan	About $0.06 to $0.10	About $290 to $475
Vapi	$0.05 hosting + each published pass-through range, lowest to highest	$0.090 to $0.143	$449 to $715
ElevenLabs Agents (Scale)	$299 plan with 3,738 minutes, 1,262 extra at $0.08, plus model and carrier at cost	About $0.10 to $0.14	About $480 to $695
Retell AI	$0.055 infrastructure + $0.015 voice + $0.048 Gemini 3.5 Flash + $0.015 US telephony. ElevenLabs voice adds $0.025	$0.133 to $0.158	$665 to $790
Bland AI (Start)	$0.14 connected minute with model and voice included. Build would cost $899 at this volume	$0.140	$700
Synthflow	$30,000 a year minimum contract, which is $2,500 a month before a single call	$0.50 floor at this volume	$2,500 minimum

Excluding Synthflow, the whole field sits inside a band of about $0.06 to $0.158 a minute, and at 5,000 minutes the gap between the cheapest and most expensive realistic setup is roughly $500 a month. That is real money, but it is smaller than most migration projects cost. So if your reason for leaving Vapi is the per-minute bill alone, check the model line first. Moving from a GPT 4.1 class model to a smaller one changes Vapi pricing by up to $0.0375 a minute without changing platform at all. If Retell AI is the main contender, the Retell AI pricing vs Vapi breakdown prices every Retell model against Vapi and finds the concurrency break-even at about 3,000 minutes a month.

The two cheapest routes, Telnyx and LiveKit, are also the ones that expect the most engineering. LiveKit in particular is a framework you write the agent in rather than a dashboard you fill in. That is where the saving comes from, and it is the part a voice AI agent developer earns back on a high volume line.

The break-even points

The fixed fees that decide which Vapi alternative is cheaper

Per-minute gaps are cents. Monthly minimums and add-ons are hundreds or thousands of dollars, and they are where the right answer flips with your volume. Each row below is simple division on published numbers.

Fixed cost	The arithmetic	Break-even volume	What it means
Vapi HIPAA, $2,000 a month	$2,000 divided by a $0.12 typical call minute	About 16,700 minutes a month	Below that, compliance costs more than every call you take. At 2,000 minutes it is $1.00 a minute on its own
Bland Build, $299 a month	$299 divided by the $0.02 saved per minute versus Start	14,950 minutes a month	About 5,000 three minute calls. Below that, stay on Start unless you need the 50 concurrent lines
Synthflow, $30,000 a year	$2,500 a month divided by Retell at $0.13 a minute	About 19,200 minutes a month	Against Telnyx at $0.057 the same floor needs about 43,900 minutes before it stops costing extra
LiveKit Scale, $500 a month	Ship costs about $0.014 a minute in session and SIP, Scale $0.003 after its fixed fee	About 45,500 minutes a month	Stay on Ship until then. Scale also cuts SIP from $0.004 to $0.003 a minute

The first row is the one that catches US healthcare buyers. A dental group or clinic running an AI receptionist on 2,000 minutes a month would pay about $240 for the calls and $2,000 for the HIPAA add-on. If protected health information is in scope, compare the signed BAA and its total cost across vendors, not the minute. Concurrency is the other quiet fixed cost: Vapi pay as you go includes 4 simultaneous calls against Retell at 20, and a front desk that gets five calls at 9am needs a fifth line.

Reasons to switch

Which Vapi alternative fits the reason you are leaving

People do not search for a Vapi alternative in the abstract. They hit one specific wall. Match the alternative to the wall rather than to a ranking.

Too many invoices

Vapi passes transcription, model, voice and carrier through separately. Bland AI bundles the model and voice into one connected minute, and Telnyx puts carrier and agent on one bill.

The bill is too high at volume

Telnyx and LiveKit Cloud publish the lowest unit costs here. The saving is real above tens of thousands of minutes and mostly disappears on a small line.

Not enough concurrent calls

Retell AI includes 20 concurrent calls before charging $8 a line. Bland Build includes 50. Vapi pay as you go includes 4 and charges $10 per extra line.

Compliance cost

Vapi HIPAA is a flat $2,000 a month add-on. ElevenLabs offers BAAs on Enterprise. Ask every vendor for the BAA, its scope and its total price in writing.

Voice quality and languages

ElevenLabs Agents is built around its own voices. Retell AI and Vapi can also use ElevenLabs voices, at $0.040 a minute on Retell.

Wanting to own the stack

LiveKit agents are code you write and can move. That is the most control on the list, and the most engineering to get a first call answered.

How to switch

What moving off Vapi actually involves

A platform change is part copy and paste and part rebuild. Knowing which part is which is the difference between a two day job and a two week one.

01

Inventory what the agent really does

List every tool call, webhook, transfer rule and fallback. Most production Vapi agents do more than the dashboard suggests, and the forgotten edge case is what breaks after cutover.

02

Separate what moves from what is rebuilt

Prompts, knowledge content and your own backend endpoints move. Assistant configuration, tool definitions, webhook payload handling and call flow logic are platform specific and get rewritten.

03

Rebuild the agent on the new platform

Recreate the tools against the new schema, reconnect the calendar or CRM, and set the same model and voice first, so any change in behavior is the platform rather than a new prompt.

04

Run both side by side on test calls

Script the twenty calls that matter most, including the awkward ones, and run them on both platforms. Compare outcomes and latency, not just transcripts.

05

Move the numbers last

Port the numbers or point them at the new platform over SIP only after the parallel run passes. Keep the old agent available for a week so rollback is a routing change.

The honest version

Where BotGigs fits, and where it does not

BotGigs is not a voice platform and does not bill you per minute. It is where you find the person who builds on one. Describe the agent once, including the platform you are on or considering, and get back a scoped hire brief with matched voice AI developers, their rates and their delivery location. Someone who has already shipped on Retell AI or LiveKit will tell you in the first call whether your migration is a rebuild or a reconfiguration.

What it will not do is make a small line worth migrating. If you take a few hundred calls a month, the difference between the cheapest and the most expensive platform above is tens of dollars, and the right move is usually to stay where you are and tune the model choice. Hiring a developer pays off when volume is high, when compliance costs are fixed, or when the agent has to do real work inside your systems.

If you are still deciding between a platform and a done for you service, compare the conversational AI companies that sell finished agents, and if you already know you want a builder, see how to hire AI engineers and what they charge on our AI developer pricing page.

Questions

Vapi alternative questions buyers actually ask

What is the best alternative to Vapi?

It depends on what you are replacing. For a lower all-in minute, Telnyx publishes $0.05 a minute including transcription and voice, and LiveKit Cloud comes to roughly $0.06 to $0.10. For a single bundled line item, Bland AI charges $0.14 a connected minute with the model included. For itemized managed pricing, Retell AI lands near $0.13 on a GPT 4.1 agent.

How much does Vapi cost per minute?

Vapi charges $0.05 a minute for hosting, but that is not the bill. Its own pricing page lists transcription at $0.0095 to $0.0099, the model at $0.0077 to $0.0452, the voice at $0.0146 to $0.0238 and telephony at $0.008 to $0.014 a minute on top. Added up, a typical agent costs about $0.09 to $0.14 a minute.

Is Vapi cheaper than Retell AI?

Slightly, on comparable settings. Built from each vendor published component rates, a GPT 4.1 class agent with a standard voice on US telephony costs about $0.09 to $0.14 a minute on Vapi and about $0.13 on Retell AI. The bigger difference is concurrency: Vapi pay as you go includes 4 concurrent calls, Retell includes 20 before charging $8 a line.

What is a good Bland AI alternative?

If you left Bland AI for cost, Telnyx and LiveKit Cloud both price the agent below its $0.12 to $0.14 connected minute. If you left for control, Vapi and Retell AI let you pick the model and voice separately, which Bland bundles. Bland Build costs $299 a month and only beats its own Start plan above 14,950 minutes a month.

Is Vapi HIPAA compliant?

Vapi sells HIPAA compliance as a published add-on at $2,000 a month. At 2,000 call minutes a month that is $1.00 a minute for compliance alone, several times the cost of the call. ElevenLabs offers BAAs on its Enterprise plan. Get the signed BAA and its scope from any vendor in writing before patient calls go live.

Do I need a developer to use Vapi?

For anything past a demo, usually yes. Vapi, Retell AI, Telnyx and LiveKit expect someone to write the prompts, define the tools the agent calls, connect your calendar or CRM, handle webhooks and test edge cases. Bland AI and ElevenLabs reduce the setup, but the integration work into your own systems still needs an engineer.

Can I switch from Vapi to another platform without rebuilding?

Partly. The prompts, knowledge content and your backend endpoints move across. The assistant configuration, tool definitions, webhook payload handling and call flow logic are platform specific and have to be rewritten. Phone numbers can be ported or pointed at the new platform over SIP. Plan a parallel run with real test calls before cutting over.

Which voice AI platform is cheapest?

On published rates, Telnyx at about $0.056 a minute with an open model it hosts, then LiveKit Cloud at roughly $0.06 to $0.10 once you add a model. Vapi, ElevenLabs, Retell AI and Bland AI all sit between about $0.09 and $0.158. Synthflow now publishes only an Enterprise tier, from $30,000 a year.