Vapi advertises voice AI at $0.05 per minute. That figure covers Vapi's orchestration layer only. To run a working voice agent on Vapi, customers also pay separately for speech-to-text, the language model, the text-to-speech voice, telephony, and concurrency lines. Most production deployments land between $0.15 and $0.40 per minute once all providers are stacked. Healthcare deployments add another $2,000 per month for HIPAA compliance.

This guide breaks down Vapi's pricing structure, the costs that sit outside the $0.05 headline, what production deployments cost at different volumes, and how Vapi compares to SquawkVoice's flat per-minute model.

TL;DR

• Vapi offers two plans: Build (pay-as-you-go, usage-based) and Scale (annual contract, custom enterprise pricing).
• Vapi's hosting fee is $0.05 per call minute and $0.005 per SMS or chat message. This covers the orchestration layer only.
• Real per-minute costs land between $0.15 and $0.40 once STT, LLM, TTS, and telephony providers are added. Industry analyses converge around $0.23 to $0.33 per minute for production deployments.
• Concurrency starts at 10 simultaneous calls. Additional lines cost $10 per line per month.
• HIPAA compliance is a $2,000 per month add-on. Zero Data Retention is another $1,000 per month add-on.
• SquawkVoice charges a flat $0.20 per minute that includes voice AI, telephony, recordings, transcripts, and post-call summaries on all plans.
• Vapi suits engineering teams that want full BYO stack control. SquawkVoice suits operations teams that want predictable costs without managing four to six provider relationships.

How Vapi Pricing Actually Works

Vapi is an orchestration layer. It coordinates the components that power a voice agent (speech recognition, language model, voice synthesis, and telephony), but it does not provide those components itself. Customers bring and pay for each piece separately.

That means Vapi's $0.05 per minute charge is one of five or six line items on the monthly bill. The full stack looks like this:

• Vapi platform fee: $0.05 per call minute, $0.005 per SMS or chat message
• Speech-to-Text (STT): billed by provider (Deepgram, AssemblyAI, others)
• Large Language Model (LLM): billed by provider (OpenAI, Anthropic, Google, others), priced per token
• Text-to-Speech (TTS): billed by provider (ElevenLabs, Cartesia, PlayHT, others)
• Telephony: billed by provider (Twilio, Telnyx, Vonage) for inbound and outbound calls, plus monthly phone number rentals
• Add-ons: HIPAA compliance, Zero Data Retention, additional concurrency

Sources: vapi.ai/pricing; Ringg AI Vapi pricing breakdown; Retell AI Vapi review; Zeeg Vapi pricing analysis.

Vapi's Two Plans: Build and Scale

Vapi simplified its plan structure to two tiers. The decision between them is less about features and more about whether the customer wants pay-as-you-go billing or an annual contract with enterprise compliance.

Build Plan (Pay-as-you-go)

Build is the default starting point. It is usage-based, has no upfront commitment, and bills only for what is used.

• Vapi hosting: $0.05 per call minute, $0.005 per chat message
• Model provider costs: passed through at cost (or free if customers bring their own API key)
• Concurrency: 10 simultaneous calls included; $10 per line per month for additional lines
• Call history retention: 14 days
• Chat history retention: 30 days
• Support: email and Discord community
• Free trial: 60 free minutes (some sources cite $10 in credits) on signup

Build is positioned for testing, development, and small-to-medium production deployments. It does not include SSO, RBAC, SOC 2, or other enterprise security features.

Scale Plan (Annual Contract)

Scale is Vapi's enterprise tier. Pricing is custom and requires sales conversations. There is no public per-minute number.

• Fixed platform fee plus committed call volume
• Volume-based per-minute pricing
• SOC 2, PCI, SSO, RBAC, data residency
• Custom data retention
• Priority access to model and voice providers
• Dedicated account team and named support engineer
• Custom support SLA and enterprise infrastructure SLA

Independent industry analyses estimate enterprise Vapi contracts typically range from $40,000 to $70,000 per year for moderate-volume deployments, though final pricing depends on committed volume and negotiation.

Sources: vapi.ai/pricing; CloudTalk Vapi pricing analysis; Emitrr Vapi pricing breakdown.

Vapi Pricing at a Glance

The Real Cost: What $0.05 Becomes in Production

The advertised $0.05 per minute covers Vapi's orchestration only. The actual cost of running a voice agent depends on which providers are stacked on top. The table below shows a typical production deployment based on independent analyses from multiple voice AI buyer-side reviews:

Sources: Ringg AI Vapi pricing analysis (April 2026); Lindy Vapi review (January 2026); Zeeg Vapi pricing breakdown (April 2026); pxlpeak Vapi pricing analysis (April 2026); Emitrr Vapi pricing breakdown (February 2026).

Industry buyer-side analyses converge on this range consistently. A simple configuration using GPT-4o mini and a basic voice sits near the low end. A higher-quality configuration with premium voices and frontier LLMs lands at the top. Most production deployments end up between $0.23 and $0.33 per minute once teams settle on configurations that work.

Costs That Are Easy to Miss

Concurrency line fees

The Build plan includes 10 simultaneous call lines. Twenty lines would cost an additional $100 per month. An outbound sales campaign running 50 concurrent calls would cost an additional $400 per month on top of usage. For high-volume operations, concurrency costs alone can match a small SaaS subscription.

HIPAA compliance at $2,000 per month

Vapi charges $2,000 per month as a flat add-on for HIPAA compliance and a Business Associate Agreement (BAA). This applies to both Build and Scale plans. For dental offices, healthcare clinics, or med spas evaluating voice AI, that represents $24,000 per year before any call traffic.

Zero Data Retention at $1,000 per month

Compliance requirements that include zero data retention add another $1,000 per month on top of HIPAA. Together, the two add-ons total $3,000 per month for regulated industries.

Telephony rates and number rentals

Twilio and Telnyx charge per-minute rates plus monthly fees for each phone number. Businesses that need local numbers across multiple area codes (common for service providers covering different regions) accumulate number rental costs quickly.

Failed and short calls still cost money

If the AI answers and the caller hangs up after three seconds, all providers are still billed for those three seconds. Spam traffic on inbound numbers can create real costs with no business value.

Development and integration time

Vapi is developer-first by design. Building a production-ready voice agent typically requires 20 to 40 hours of engineering work for prompt design, integration setup, testing, and tuning. That translates to $2,000 to $4,000 in labor before the first paying call. Non-technical teams generally cannot deploy Vapi without engineering support.

Latency stacking

Because Vapi chains four to five different APIs together, each provider hop adds milliseconds of latency. Across a five-minute call, this can compound into noticeable conversational delay. This is a quality cost rather than a billing cost, but it affects whether the deployment works.

Sources: vapi.ai/pricing; Retell AI Vapi review; Ringg AI Vapi pricing analysis; VoiceFleet Vapi review (March 2026).

What Vapi Actually Costs at Different Volumes

The estimates below use a mid-range stack: Deepgram for STT, GPT-4o for LLM, ElevenLabs Turbo for TTS, and Twilio for telephony. Real costs vary based on provider choices, prompt length, and call concurrency.

Estimates exclude concurrency fees, phone number rentals, and development time. HIPAA add-on applies only to regulated industries.

Two patterns stand out. First, the $0.05 per minute headline becomes $0.25 per minute or higher in any working deployment, roughly a 5x multiplier. Second, HIPAA is a step change: a small dental practice running 1,000 minutes a month pays more for the compliance add-on than for actual call traffic.

Vapi vs SquawkVoice: Pricing Comparison

Vapi and SquawkVoice solve similar problems but take different approaches to pricing. Vapi is build-your-own-stack: low base fee, multiple vendors, maximum flexibility. SquawkVoice is integrated: one flat per-minute price that includes the full stack.

Cost Comparison at Different Volumes

The table below shows the same volume tiers with SquawkVoice's flat $0.20 per minute rate added. This compares call-traffic costs only and excludes vendor management time, development hours, and concurrency fees.

At every volume tier shown, SquawkVoice's flat rate sits below a realistic Vapi production stack on raw per-minute cost. The gap widens once HIPAA, concurrency, and development time are factored in. For healthcare practices that need HIPAA, SquawkVoice avoids the $24,000 per year compliance add-on entirely.

Vapi can come in lower for deeply technical teams that build a minimal stack with low-cost providers (GPT-4o mini, basic voices, no premium features). For those teams, $0.05 to $0.15 per minute is achievable. That configuration also produces the latency and voice-quality issues Vapi users consistently mention in reviews.

Where Each Tool Wins

Choose Vapi if:

• You have engineering resources and want full control over every layer of the voice stack
• Voice AI is a core product feature you are building, not an operational tool you are buying
• You want to swap providers (LLMs, voices) freely as the market evolves
• Your use case requires deep customization that no integrated platform supports
• You can manage multiple vendor contracts, invoices, and support relationships

Choose SquawkVoice if:

• You want predictable monthly costs without modeling five separate provider invoices
• You run a dental office, healthcare clinic, med spa, HVAC business, roofing company, or other service-based business that needs answering and booking automation
• You need HIPAA without paying $2,000 per month on top of usage
• Your team does not have engineering capacity to build, test, and maintain a custom voice stack
• You want call recordings, transcripts, and post-call summaries included rather than configured
• You need a mobile app for fast SMB self-service setup

When Vapi's Pricing Makes Sense

Vapi's pricing model fits specific scenarios where build-your-own-stack delivers real value:

• Building a voice product to sell: If voice is core to the offering, paying for orchestration only and managing the rest of the stack gives teams flexibility and margin.
• Workflows requiring proprietary models: Some use cases need specific LLMs, fine-tuned voices, or custom STT engines that integrated platforms do not support.
• Engineering-heavy teams: Companies with dedicated developer resources can extract value from Vapi's API-first design and avoid the markup that integrated platforms charge.
• High-volume technical operations: At very high volumes, optimizing each provider individually can yield meaningful savings, given the engineering capacity to do that optimization.

Vapi's pricing model gets harder to navigate when teams without engineering resources discover that the $0.05 per minute headline does not reflect operational reality. By the time contracts are signed with four providers, HIPAA is paid for, and 30 hours have been spent building the agent, total cost of ownership is higher than a flat per-minute integrated platform, and the resulting system has more moving parts.

Frequently Asked Questions

Is Vapi really $0.05 per minute?

The $0.05 per minute charge is Vapi's platform fee only. It does not include the language model, speech-to-text, text-to-speech, or telephony providers required to run a working voice agent. Most production deployments cost $0.15 to $0.40 per minute once all providers are stacked, with $0.23 to $0.33 being the typical range.

Does Vapi include telephony?

No. Vapi orchestrates the AI conversation layer, but actual phone calls require a separate telephony provider like Twilio, Telnyx, or Vonage. Customers sign a contract with the telephony provider directly, pay their per-minute rates, and pay monthly fees for each phone number.

How much does HIPAA compliance cost on Vapi?

Vapi charges $2,000 per month as a flat add-on for HIPAA compliance and a Business Associate Agreement. This applies to both the Build (pay-as-you-go) and Scale (enterprise) plans. That represents $24,000 per year before any call traffic, a significant consideration for healthcare practices evaluating voice AI.

What does an enterprise Vapi contract cost?

Vapi's Scale plan requires a custom quote with an annual contract and committed call volume. Independent industry estimates place typical enterprise budgets between $40,000 and $70,000 per year for moderate-volume deployments, though final pricing depends on negotiation and volume commitments.

Can I get fixed monthly pricing on Vapi?

Not on the standard Build plan. Vapi's model is fully usage-based, and monthly cost varies with call volume, model selection, and provider choices. Fixed pricing is only available through Scale plan negotiations or by building custom usage caps on top of the API.

How does Vapi compare to SquawkVoice on price?

SquawkVoice charges a flat $0.20 per minute that includes voice AI, telephony, call recordings, transcripts, and post-call summaries, with no separate vendor contracts. Vapi's headline rate is lower at $0.05 per minute, but real-world deployments typically land at $0.23 to $0.33 per minute once the full stack is assembled. SquawkVoice is generally cheaper at every volume tier and significantly cheaper for healthcare practices that need HIPAA, since SquawkVoice does not charge a $2,000 per month compliance add-on.

Does Vapi have caller or concurrency limits?

The Build plan includes 10 simultaneous call lines. Additional concurrency costs $10 per line per month. High-volume operations or outbound campaigns running dozens of concurrent calls add meaningful monthly costs on top of per-minute usage.

What free trial does Vapi offer?

Vapi offers either 60 free minutes or $10 in starter credits on signup, depending on the source. There is no ongoing free tier. Once trial credits run out, all usage is billed.

Is Vapi suitable for non-technical teams?

Vapi is developer-first by design. Production deployments typically require 20 to 40 hours of engineering work for prompt design, integration setup, and testing. Non-technical teams generally cannot deploy Vapi successfully without engineering support. For business owners or operations leaders who want voice AI without the build, integrated platforms like SquawkVoice are a better fit.

The Bottom Line on Vapi Pricing

Vapi's $0.05 per minute headline is the cheapest entry point in voice AI infrastructure. Real deployments cost five times the advertised rate once the required providers are stacked. Healthcare deployments add another $24,000 per year for compliance.

For engineering teams building voice products, that flexibility is valuable. For operations leaders who want every call answered, appointments booked, and customers routed without managing four vendors and writing custom code, an integrated platform with predictable flat pricing (like SquawkVoice) usually wins on total cost of ownership and time to value.

The right question is whether the engineering capacity exists to make the $0.05 per minute rate matter in production.

Want to see how SquawkVoice handles voice AI without the vendor-stacking math? Start your free trial with 50 free calls included, no credit card required.