AI voice agent pricing comes in two main shapes. Per-minute platforms bill every connected minute, so the bill follows talk time. Per-bot pricing, which is how VoiceOI bills, charges $25 to $155 per bot per month at list price, so the bill follows how many calls you need live at once, with volume discounts up to 60%. Carrier minutes are billed separately either way.
This guide shows the cost stack behind per-minute prices, the rates five vendors publish, VoiceOI's tiers and discounts, low, medium and high volume scenarios, and a hidden-fee checklist. Last updated October 1, 2026.
Five pricing models cover almost every product marketed as an AI voice agent or AI dialer. They differ in what you are charged for, so they also differ in what makes the bill grow.
| Model | How it bills | Examples from vendor pages | What scales the bill |
|---|---|---|---|
| Bundled per-minute | One rate per connected minute that includes several layers | Bland AI: $0.14/min on Start, $0.12/min plus $299/month on Build, with LLM, transcription and voices in the rate | Talk time |
| Per-minute stack | Platform fee per minute plus separately priced model, voice and telephony costs | Vapi: $0.05/min hosting plus pass-through model costs. Retell AI: $0.055/min infrastructure plus LLM, voice and telephony. Telnyx: $0.05/min voice engine plus LLM tokens | Talk time, model choice, voice choice |
| Enterprise contract | Annual contract scoped to volume and requirements | Synthflow: enterprise contracts start at $30,000 annually. Bland AI: custom Enterprise plan | Contract size and scope |
| Per seat | Per user per month for software a human rep operates | Dialer vendors, covered in AI dialer cost: per-seat vs per-bot pricing | Rep headcount and dialing mode |
| Flat per bot | Prepaid price per bot per month; one bot is one concurrent call | VoiceOI: $25 to $155 per bot per month at list price, volume discounts up to 60% | Concurrent bots, not minutes |
Sources: Bland AI's pricing page, Vapi's pricing page, Retell AI's pricing page, Telnyx's voice AI agent pricing page, Synthflow's pricing page, reviewed October 2026. Competitor pricing changes often; check each vendor's page before you buy.
A connected AI call uses up to five paid layers: speech-to-text, a language model, text-to-speech, telephony, and the platform that orchestrates them. A headline per-minute rate may include some of them or all of them, which is why two vendors with the same headline can produce very different bills. The table lists what each vendor publishes for each layer.
| Layer | What vendors publish | Source |
|---|---|---|
| Platform or voice infrastructure | Retell AI: $0.055/min voice infrastructure. Vapi: $0.05/min hosting. Telnyx: $0.05/min voice engine (orchestration, speech-to-text, text-to-speech), with a $0.002/min platform fee listed separately. | Retell AI's pricing page; Vapi's pricing page; Telnyx's voice AI agent pricing page |
| Speech-to-text | Vapi passes transcription through at $0.0095 to $0.0099/min. Bland AI lists real-time transcription inside its per-minute rate. Telnyx bundles speech-to-text in its voice engine. | Vapi's pricing page; Bland AI's pricing page; Telnyx's voice AI agent pricing page |
| Language model (LLM) | The most variable layer. Retell AI lists LLMs from $0.0016/min (GPT 5 nano) to $0.32/min (GPT 6 Astra) on its standard tier, with a fast tier priced higher. Vapi: $0.0077 to $0.0452/min. Telnyx: $0.0064/min in its example, billed per token. Bland AI lists its LLM inside its rate. | Retell AI's pricing page; Vapi's pricing page; Telnyx's voice AI agent pricing page; Bland AI's pricing page |
| Text-to-speech | Retell AI: $0.015/min for most voices, $0.040/min for ElevenLabs voices. Vapi: $0.0146 to $0.0238/min for ElevenLabs. Bland AI lists premium voices inside its rate. | Retell AI's pricing page; Vapi's pricing page; Bland AI's pricing page |
| Telephony | Retell AI: $0.015/min default with US Twilio. Vapi lists Twilio outbound at $0.014/min and Telnyx at $0.0055/min. Telnyx lists outbound SIP trunking from $0.005/min. Bland AI bills telephony separately and supports bring-your-own telephony. | Retell AI's pricing page; Vapi's pricing page; Telnyx's voice AI agent pricing page; Bland AI's pricing page |
Add the layers and you can land well above the lowest number on a pricing page. Retell AI's own page states an overall range of $0.07 to $0.31 per minute for AI voice agents, yet its component table can add up to more than that for premium models. Telnyx's example estimate of about $0.06 per minute includes local inbound US telephony, so it is not directly comparable to an outbound rate. To test any of these rates against a per-bot bill, use the AI calling cost per minute calculator. VoiceOI's speech, voice and language model run on its own infrastructure by default and are covered by the per-bot price rather than listed as separate per-minute lines.
Sources: Bland AI's pricing page, Vapi's pricing page, Retell AI's pricing page, Telnyx's voice AI agent pricing page, reviewed October 2026. Competitor pricing changes often; check each vendor's page before you buy.
| Vendor | Pricing on its own page | Limits and extras it lists | Source |
|---|---|---|---|
| Bland AI | Start: $0.14/min, no platform fee. Build: $0.12/min plus $299/month. Enterprise: custom. LLM, transcription and voices are listed in the rate. | Transfer time $0.05/min (Start) and $0.04/min (Build). Start lists 10 concurrent calls and 100 calls a day; Build lists 50 concurrent calls and 2,000 calls a day. Telephony billed separately; bring-your-own telephony supported. | Bland AI's pricing page |
| Vapi | $0.05/min hosting, plus transcription, language model and voice costs passed through at cost, plus telephony by provider. | Core plan $29/month with 10 concurrent calls; Pro from $999/month with 30. Extra concurrency $10 per line per month. Its calculator estimates $82 to $129 a month for 1,000 minutes with no success package. | Vapi's pricing page |
| Retell AI | Voice infrastructure $0.055/min; LLM ($0.0016 to $0.32/min on the standard tier) and voice ($0.015 or $0.040/min) priced by choice; telephony $0.015/min by default. Stated overall range $0.07 to $0.31/min. | 20 concurrent calls included, then $8 per concurrency per month. Phone number rental $2/month. $10 in free signup credits listed. | Retell AI's pricing page |
| Telnyx | $0.05/min voice engine. Its example estimate of about $0.06/min adds $0.0032/min of local inbound US telephony and $0.0064/min of LLM tokens; a $0.002/min platform fee and outbound SIP trunking (from $0.005/min) are listed separately. | 500 concurrent calls included on pay-as-you-go; no monthly minimum listed. | Telnyx's voice AI agent pricing page |
| Synthflow | Enterprise contracts start at $30,000 annually; final pricing is scoped around call volume, concurrency, telephony setup, integrations, security needs and launch support. | The page lists no per-minute rate. | Synthflow's pricing page |
Sources: Bland AI's pricing page, Vapi's pricing page, Retell AI's pricing page, Telnyx's voice AI agent pricing page, Synthflow's pricing page, reviewed October 2026. Competitor pricing changes often; check each vendor's page before you buy. Free credits, discounts and plan names can change at any time.
Head-to-head comparisons: VoiceOI vs Bland AI, VoiceOI vs Vapi and Retell AI and VoiceOI vs Synthflow.
VoiceOI is prepaid per bot per month. Prices start at $25 for a fronter bot, $50 for a generative lead-gen bot and $100 for a closer bot, at list price. A bot is one concurrent call slot. Fronter and closer bots transfer qualified calls to a human closer; lead-gen tiers do not include live transfers. Hosted SIP tiers use a dedicated switch; VICIdial and Asterisk tiers connect to your existing dialer.
| Tier | Connection | List price | Price floor | Includes |
|---|---|---|---|---|
| VICIdial Simple Fronter | VICIdial / Asterisk | $25 | $18 | Basic filtering, No Closer comments, Live Transfers, Connects to VICIdial / Asterisk |
| VICIdial Good Fronter | VICIdial / Asterisk | $35 | $22 | Verifies Age/Zip, Leaves comments for Closer, Live Transfers, Connects to VICIdial / Asterisk |
| SIP Simple Fronter | Hosted SIP | $45 | $28 | Basic filtering, No Closer comments, Live Transfers, Dedicated Switch |
| SIP Good Fronter | Hosted SIP | $55 | $34 | Verifies Age/Zip, Leaves comments for Closer, Live Transfers, Dedicated Switch |
| VICIdial Simple LG | VICIdial / Asterisk | $50 | $30 | Generative AI, No Google Sheets export, No Live Transfers, Connects to VICIdial / Asterisk |
| VICIdial Pro LG | VICIdial / Asterisk | $65 | $40 | Generative AI, Google Sheets export, No Live Transfers, Connects to VICIdial / Asterisk |
| SIP Simple LG | Hosted SIP | $70 | $45 | Generative AI, No Google Sheets export, No Live Transfers, Dedicated Switch |
| SIP Pro LG | Hosted SIP | $85 | $55 | Generative AI, Google Sheets export, No Live Transfers, Dedicated Switch |
| VICIdial Standard Closer | VICIdial / Asterisk | $100 | $65 | Live Transfers, Connects to VICIdial / Asterisk |
| VICIdial Ultimate Closer | VICIdial / Asterisk | $135 | $85 | Complex branching, Live Transfers, Connects to VICIdial / Asterisk |
| SIP Standard Closer | Hosted SIP | $120 | $75 | Live Transfers, Call Termination on Transfer, Dedicated Switch |
| SIP Ultimate Closer | Hosted SIP | $155 | $95 | Complex branching, Live Transfers, Call Termination on Transfer, Dedicated Switch |
| Bots (concurrent seats) | Discount off list price |
|---|---|
| 1 to 9 | None (standard rate) |
| 10 to 49 | 15% off |
| 50 to 99 | 30% off |
| 100 to 249 | 45% off |
| 250 or more | 60% off |
Discounts apply to every tier and no tier drops below its floor. The Standard Voice Engine is listed as “Included at no extra cost” and the Advanced Voice Engine as “2 Workers Free (then $15/worker)”. Full details are on the VoiceOI pricing page.
These monthly scenarios come from stated assumptions, not customer data. Minutes are connected minutes. VoiceOI is priced on the SIP Good Fronter tier ($55 list per bot) with the volume discount for each fleet size. The per-minute column spans the lowest and highest cost across Bland AI, Vapi, Retell AI and Telnyx at the usage rates above, before carrier charges. Carrier minutes use $0.014per minute, the Twilio outbound rate on Vapi's pricing page, and apply to every platform.
| Scenario | Assumptions | VoiceOI per bot | Per-minute platforms, lowest to highest | Carrier minutes |
|---|---|---|---|---|
| Low volume | 5,000 connected minutes, 2 bots | $110$0.022/min effective; Standard Rate (1-9 Seats) | $292 to $2,075 | $70 |
| Medium volume | 50,000 connected minutes, 10 bots | $468$0.0094/min effective; 15% Off (10-49 Seats) | $2,920 to $20,750 | $700 |
| High volume | 300,000 connected minutes, 50 bots | $1,925$0.0064/min effective; 30% Off (50-99 Seats) | $17,520 to $124,500 | $4,200 |
Read these as shapes, not promises. The per-minute range is wide because the language model and voice you pick move the rate, and the highest figures use the highest-priced LLM and voice options Retell AI lists on its standard tier. The per-bot figure is fixed by fleet size, but it only turns into a low effective rate if the bots are connected for a good part of the day: here each bot carries between 2,500 and 6,000 connected minutes a month, and these effective rates assume your plan has no minute cap per bot; confirm the terms before you buy. Change the inputs in the cost calculator to match your own volumes. For the plain-language version of this trade-off, read Per-Minute vs Per-Bot Pricing for Voice AI.
A fronter is the first-contact caller who qualifies a lead before a closer takes it. To compare a bot with a human fronter, use your own numbers: your fully loaded monthly cost per fronter seat (pay, benefits, software, supervision), the qualified transfers a fronter produces per month, and the same two figures for a bot (the per-bot price plus carrier minutes, and the transfers it produces). The measure that matters is cost per qualified transfer, not cost per seat. The two are not one-for-one: a bot holds one call at a time and does not replace the closers who take the transfer.
Seat-based dialers, which are a different cost from either, are covered in AI dialer cost: per-seat vs per-bot pricing. If you are choosing between platforms, see Bland AI alternatives and Retell AI alternatives.
Per minute: connected minutes x the vendor's rate. Per bot: number of bots x the per-bot price. Do both in the cost calculator, then confirm current VoiceOI tiers on the VoiceOI pricing page or talk to sales. VoiceOI publishes this page and has an interest in the comparison, so check each figure against the linked vendor pages. Competitor information changes often. Last reviewed: October 2026.
How to compare per-bot and per-minute voice AI pricing.
On per-minute platforms, the published platform layer starts around $0.05 per minute (Vapi $0.05/min hosting, Retell AI $0.055/min, Telnyx $0.05/min voice engine), with the language model, voices and telephony billed on top. Bland AI lists $0.14/min on Start and $0.12/min plus $299 a month on Build. VoiceOI charges per bot instead: $25 to $155 per bot per month at list price, with no per-minute AI fee. Your carrier bills VoiceOI call minutes.
Per-minute platforms charge for every minute a call is connected, so the bill rises with call volume and talk time. Per-bot pricing charges for concurrent capacity instead, so the monthly AI cost is not billed per minute. Which is cheaper depends on how many minutes each bot actually handles and on the per-minute rate you compare against, so run the numbers with your own dialer talk-time data. Effective per-minute figures for bots assume your plan has no minute cap per bot; confirm the terms before you buy.
Include carrier and telephony charges, the AI platform fee, any separately billed speech recognition, voice synthesis or language-model usage, transfer fees, setup fees and minimums, concurrency and daily-call limits, and the cost of the human closers who receive transferred calls. Carrier charges are billed by your carrier and are separate from VoiceOI's per-bot price.
Five layers: speech-to-text, a language model, text-to-speech, telephony, and the platform that orchestrates them. Some vendors bundle several layers into one rate (Bland AI lists LLM, transcription and voices inside its per-minute price), and others list each layer separately (Vapi and Retell AI), so compare the all-in rate rather than the headline. Rates are from each vendor's own pricing page, reviewed October 2026.
No. The AI side is a prepaid monthly price for each bot ($25 to $155 at list price, set by bot type and connection, with up to 60% off at volume). Only your carrier bills by the minute.
It depends on how much each bot talks. When bots are connected for many hours a day, a flat per-bot price spreads over more minutes and can cost less per minute. When bots are lightly used, a per-minute rate can cost less. The platforms also differ in what they do: VoiceOI focuses on outbound fronting for call centers with live transfer to human closers, VICIdial and SIP, and does not offer a public API or inbound calling. Use the calculator with your own volumes.
No. Carrier minutes are billed by your carrier. With VoiceOI you bring your own SIP carrier, such as a Twilio or Telnyx trunk, or use the carriers behind your VICIdial setup.