What a voice AI minute actually costs: component economics from published 2026 prices

Every AI phone call has five cost components: speech-to-text (STT), language-model inference (LLM), text-to-speech (TTS), telephony, and orchestration or hosting. Published list prices exist for each. Assembling them explains how lean DIY claims can approach half a cent per minute, why a practical component floor is closer to 1–4¢ per connected minute, and why managed offerings commonly reach roughly $0.05–$0.31 per minute depending on what is included. Examples include Telnyx pricing, Retell pricing, and AssemblyAI pricing, accessed 2026-09-01.

The hard part is normalization. Speech services may bill audio time, open-session time, characters, tokens, or wall-clock call time. The calculations below convert those units into one connected-minute view. Every conversion and assembled total is an illustrative planning estimate, not a quote. The named component vendors are market reference points, not statements about which suppliers ThunderPhone uses.

Conversion assumptions come first

A useful estimate must state how much speech and model activity occurs during a minute:

These are workload assumptions, not universal constants. Long system prompts, retrieved context, tool results, interruptions, silence, and the agent's speaking share can move the result. For TTS, 400–800 generated characters per connected minute is a reasonable planning band under Inworld's published conversational model, accessed 2026-09-01. For LLMs, the two token assumptions show why prompt design matters as much as the nominal model price.

Speech-to-text: streamed-minute pricing

The AssemblyAI conversion is illustrative and not a quote: $0.15 ÷ 60 = $0.0025/minute. That makes roughly $0.0025–$0.006 a useful current list-price band for the streamed options shown, while Deepgram's regular $0.0077 rate sits above it. AssemblyAI also bills session duration, meaning the time its WebSocket remains open, rather than audio duration. Silence can therefore remain billable. AssemblyAI pricing, accessed 2026-09-01.

LLM inference: usually fractions of a cent, but context changes it

Illustrative planning estimates, not quotes, show the sensitivity:

• gpt-4o-mini: (2,340 × $0.15 + 100 × $0.60) ÷ 1,000,000 = $0.000411/minute
• Gemini 2.5 Flash: (2,340 × $0.30 + 100 × $2.50) ÷ 1,000,000 = $0.000952/minute
• Llama 3.3 70B: (2,340 × $1.04 + 100 × $1.04) ÷ 1,000,000 = $0.0025376/minute
• Llama 3.3 70B with the heavier prompt assumption: (6,000 × $1.04 + 100 × $1.04) ÷ 1,000,000 = $0.006344/minute

A rounded $0.001–$0.01 per connected minute is therefore a practical planning band for cheap-to-mid configurations. It is not a universal model price. Re-sending conversation history, tools, retrieval, and larger output can raise it.

Text-to-speech: price the agent's speaking share

For the $4–$30 market band, the illustrative conversion is not a quote: $4 ÷ 1,000,000 × 1,000 = $0.004 and $30 ÷ 1,000,000 × 1,000 = $0.03 per generated speaking minute. If the agent speaks half the call, the connected-minute contribution becomes 50% × $0.004 = $0.002 to 50% × $0.03 = $0.015. ElevenLabs' listed rate is outside that band, which is why voice selection can materially change the stack total.

Telephony: published US rates

Twilio lists US local inbound at $0.0085/minute and outbound at $0.0140/minute. Twilio US voice pricing, accessed 2026-09-01. Telnyx lists inbound local from $0.0032/minute and outbound local from $0.005/minute. Telnyx SIP pricing, accessed 2026-09-01. Daily lists PSTN dial-in or dial-out at $0.018/minute for Pipecat Cloud. Daily Pipecat Cloud pricing, accessed 2026-09-01.

Direction, number type, destination, and carrier setup determine the actual rate. Number rental and recording can add separate fixed or usage charges, so a phone-call estimate should not silently treat telephony as included.

Orchestration and media: hosted or self-run

LiveKit Cloud lists agent sessions at $0.01/minute, while its telephony is another $0.01/minute. LiveKit pricing, accessed 2026-09-01. Daily lists Pipecat Cloud agent-1x hosting at $0.01/minute active, with free transport for one-to-one voice and separately priced PSTN. Daily Pipecat Cloud pricing, accessed 2026-09-01.

Self-hosting removes the hosted vendor's list-price line, but it does not make orchestration free. Servers, media routing, scaling, observability, deployment, and on-call ownership still need a workload-specific budget.

Assemble the component floor

A lean self-managed subtotal before assigning any server or operations cost is, illustratively and not as a quote:

$0.0025 STT + $0.001 LLM + $0.002 TTS + $0.0032 telephony = $0.0087/minute.

Add published hosted orchestration and the same estimate becomes $0.0087 + $0.0100 = $0.0187/minute. A higher lean hosted mix computes to $0.006 STT + $0.006 LLM + $0.015 TTS + $0.005 telephony + $0.010 orchestration = $0.042/minute. The inputs draw from AssemblyAI pricing, Together AI pricing, Google Cloud TTS pricing, Telnyx SIP pricing, and LiveKit pricing, accessed 2026-09-01.

That supports a practical shorthand of roughly 1–4¢ per connected minute for lean raw components, depending especially on whether orchestration is self-run and how its cost is allocated. With hosted orchestration, the displayed examples are closer to 1.9–4.2¢. At 10,000 connected minutes, the illustrative arithmetic, not a quote, is 10,000 × $0.0187 = $187 to 10,000 × $0.042 = $420, before the costs in the next section.

Two external cross-checks show why no single floor is universal. The hypercheap-voiceAI repository claims $0.28/hour and labels that $0.0046/minute; the direct illustrative conversion is $0.28 ÷ 60 ≈ $0.00467/minute, not a quote. hypercheap-voiceAI, accessed 2026-09-01. Inworld models complete DIY stacks at $0.007–$0.091 per conversation minute. Inworld's voice-agent cost model, accessed 2026-09-01. As a bundled comparator, AssemblyAI lists its Voice Agent API at $4.50/hour, whose illustrative conversion is $4.50 ÷ 60 = $0.075/minute. AssemblyAI pricing, accessed 2026-09-01.

What the component floor leaves out

The floor is a vendor-bill estimate, not total cost of ownership. A production system also needs engineering to join streaming audio, turn detection, prompts, tools, and failure handling. It needs monitoring that can distinguish carrier trouble from speech, model, or application trouble. Regression suites and recorded-call review consume time and often paid usage.

Concurrency adds another gap. Average minutes do not reveal the capacity required at a noon traffic spike. Redundancy, regional failover, rate-limit headroom, and incident response all cost something even when idle. Compliance work includes contracts, access controls, retention, audit trails, and correct workflow configuration. Support and ongoing tuning remain human work.

Quality is not a monotonic function of price: a cheaper component is not automatically worse, and an expensive one is not automatically right for a workload. But the lowest-cost combination can lose accuracy, naturalness, interruption handling, language coverage, or reliability that the application needs. Test complete configurations on representative calls. A managed rate above the component floor can still produce a lower total cost when it replaces enough integration and operating work.

How managed platforms map onto these components

Retell exposes component-style pricing. Vapi charges hosting while model and speech costs pass through. Bland bundles the model, STT, and TTS into its talk-minute tiers. Telnyx bundles orchestration, STT, and TTS, with LLM and telephony on top. ThunderPhone bundles speech recognition, the language model, and voice into one engine rate; its structure is detailed in ThunderPhone pricing. These are different billing boundaries, so headline rates should not be ranked until the same components are included.

ThunderPhone Spark is 2¢/minute for those engine components, with no subscription, seat fee, or platform fee. In plain illustrative arithmetic, not a quote, 10,000 × $0.02 = $200. That sits inside the assembled component-floor range without requiring the buyer to assemble those engine components. A selected premium voice or language path can add up to 3¢/minute, or illustratively 10,000 × $0.03 = $300 at that volume, and telephony depends on setup. See the cheapest AI voice-agent scenario and rerun the arithmetic for your own call length, language, and configuration.

FAQ

What does it cost to run a voice AI agent per minute?

A lean raw-component plan is roughly 1–4¢ per connected minute, while complete managed pricing can be higher. The exact answer depends on token throughput, agent speaking share, telephony direction, orchestration, and features. The hosted examples assembled above calculate to about 1.9–4.2¢ per minute before engineering and operational costs. These are illustrative planning estimates, not quotes.

What's the cheapest possible voice AI stack?

The lowest published third-party claim reviewed here is about half a cent per minute, but that is not a safe universal budget. The hypercheap-voiceAI repository claims $0.0046/minute, while Inworld's broader DIY model spans $0.007–$0.091/minute. hypercheap-voiceAI and Inworld's cost model, accessed 2026-09-01. A buyer should price engineering, hosting, redundancy, monitoring, and support alongside API spend.

Why do platforms charge more than the component floor?

Because a production voice agent is more than five API invoices. A platform can supply integration, media handling, deployment, scaling, observability, testing, billing, and support. Whether the premium is economical depends on which responsibilities it actually removes for your team.

Do cheap components sound worse?

Not necessarily, but price alone does not predict production quality. Lower-cost models or voices may be sufficient for short, structured workflows. Other calls need stronger recognition, reasoning, naturalness, interruption handling, or language coverage. Evaluate the whole configuration with real call conditions rather than choosing each component independently by its lowest list price.

Sources and freshness

All third-party list prices and usage assumptions on this page were checked against the linked vendor pages on September 1, 2026. Promotional prices are labeled, and conversions rest on the stated speech-share, character, and token assumptions. Every conversion and worked total is an illustrative planning estimate, not a quote. Vendor pricing changes frequently, so verify current pages and rerun the arithmetic for your actual configuration.

Video: the same comparison, with every pricing page on screen

The embedded video above walks through each vendor's public pricing page (read on 10 September 2026) and adds up what one live minute costs. Full transcript:

0:00 The question and the method

Voice AI pricing pages all look different, so let's put the platforms side by side and work out what one minute of a live phone call actually costs on each of them. The method is simple. For every platform we open its public pricing page, as of September tenth, twenty twenty-six, and add up the platform fee plus the three models every call needs: speech to text, the language model, and text to speech. Telephony, monthly fees, and compliance add-ons get noted separately. We are comparing price only. Whether the calls actually work is a separate question, and a separate video.

0:35 Vapi: 5¢ platform fee, models billed on top

First, Vapi. The Build plan is usage based, and the line that matters is here: Vapi hosting cost, five cents per minute, and it explicitly excludes model provider costs. Speech to text, the language model, and text to speech are billed at cost on top, or you bring your own API keys and pay those vendors directly. Concurrency is ten calls included, then ten dollars per line per month. HIPAA is a two thousand dollar a month add-on, and zero data retention is a thousand a month. So the five cents is the floor, not the price of a minute. A typical stack with models and telephony lands around seven cents and up.

1:11 Retell: 7¢ to 31¢, their own example is 11.5¢

Retell publishes a range: seven to thirty-one cents per minute for voice agents, pay as you go, with ten dollars of free credit. Their own cost estimator shows where a minute lands. With the default choices, the language model is four and a half cents, Retell's voice infrastructure is five and a half cents, text to speech is a cent and a half, and telephony is zero only because it assumes you bring your own. That adds up to eleven and a half cents per minute before the phone line.

1:38 Bland: 14¢, or 12¢ plus $299 a month

Bland goes the other way and bundles the models. The Start plan is fourteen cents per minute with no platform fee, but it is capped at one hundred calls a day. The Build plan drops to twelve cents per minute and adds a two hundred ninety-nine dollar monthly platform fee. Transfer time bills separately at four or five cents.

1:56 Telnyx: 5¢ engine, LLM and telephony on top

Telnyx says five cents per minute, no fine print, and it is the most complete of the itemized offers: orchestration, speech to text, and text to speech are all inside the five cents. The language model is an add-on, and the headline figure, about four tenths of a cent per minute, is for Kimi, an open-source model on Telnyx's own GPUs, not a frontier model. Frontier models from OpenAI or Anthropic are billed at those vendors' rates, so a production agent on a frontier model costs more. Telephony is carrier cost, from about a third of a cent. Their own estimator, with the open-source model, lands at about five point six cents per minute including telephony. Volume discounts start at a five hundred dollar monthly commitment.

2:41 Synthflow: enterprise only, from $30,000 a year

Synthflow no longer publishes a per-minute rate at all. The pricing page has one plan, enterprise, with contracts starting at thirty thousand dollars a year and final pricing scoped to volume. Fine for a large contact center, out of reach for a self-serve team.

2:57 ThunderPhone: 2¢, 5¢, 9¢ all in, no platform fee

Now ThunderPhone. Three tiers, each one flat rate with the models included: Spark at two cents per minute, Bolt at five, Storm at nine. Models included means all of them: ThunderPhone runs several models that check each other rather than one transcription model, one language model, and one voice, and that whole stack is inside the rate. There is no platform fee, no monthly fee, and no per-line concurrency charge. Telephony is either your own SIP trunk, connected for free, or a US phone number from ThunderPhone at a dollar a month plus one cent per minute. Hold time bills at two cents a minute on every tier, and the optional add-ons are listed with their prices on the page. And for high-volume enterprise use, ThunderPhone goes lower still: one cent a minute or less under an enterprise agreement, which is what makes high-volume, low-margin outbound work.

3:40 Scoreboard: cost per live minute

Here is the scoreboard, platform plus models, per live minute, telephony noted separately. ThunderPhone Spark, two cents. Bolt, five. Storm, nine, with all the models included. Telnyx, about five point four cents once you add the open-source language model. Vapi, five cents plus whatever the models cost. Retell, seven to thirty-one cents, eleven and a half in their own example. Bland, fourteen cents, or twelve plus the monthly fee. Synthflow, enterprise contracts only. Even ThunderPhone's most capable tier, Storm, is at or below most competitors' entry rates, and the cheapest tier is less than half of the next cheapest platform.

4:18 What a 10,000-minute month costs

In dollars, for a team doing ten thousand live minutes a month: two hundred dollars on ThunderPhone Spark, five hundred on Bolt, nine hundred on Storm. Around five hundred forty on Telnyx. Five hundred on Vapi before the model bills. Eleven hundred fifty on Retell using their example. Fourteen hundred on Bland Start, if the daily cap even allows it, or fourteen ninety-nine on Bland Build.

4:41 Why ThunderPhone can be the cheapest

Why the gap? ThunderPhone does not charge a platform fee on top of model costs, and it tunes its stack for performance per dollar. All of these numbers come from the vendors' own public pricing pages, linked in the description with the date they were read. Prices change, so check the links. As of today, ThunderPhone is the cheapest voice AI platform, from two cents a minute, all in.