Lexicost
AI voice agents / Vendors

Vapi

In its own words: “Your agents, your rules. Build, deploy, and improve voice agents as fast as you ship.”

10 of 11 key facts published, pages read .

Developer platforms. Hosted, self-serve services that run the whole call: phone numbers, speech, the AI model and call control, usually priced per minute.

What Vapi publishes, in short. Platform fee per minute: (No Success Package (Usage only)). What that fee includes: (All packages). Monthly plan fee: (Core). How the AI model is charged: (All packages). Every figure below opens Vapi's own words and the date we read them.

Work out a cost with these figures: billed minutes: how vendors round a call, concurrency limits for AI voice agents, cost per call: AI agent or a person, aI voice agent cost per minute, openAI Realtime API cost per minute. Compare with other vendors in the side by side table.

Price

Platform fee per minuteNo Success Package (Usage only)
What that fee includesAll packages
Monthly plan feeCore
Minimum commitmentPro
How the AI model is chargedAll packages
How the AI model is charged
How the AI model is charged
How the AI model is charged
How the AI model is charged
US phone minutesVapi Telephony / SIP (free Vapi numbers)
PlanUsage only (No Success Package)
PlanCore
PlanPro
PlanPremier
AI model rate per minute
AI model rate per minute
AI model rate per minute
Speech-to-text rate
Speech-to-text rate
Speech-to-text rate
Voice (text-to-speech) rate
Voice (text-to-speech) rate
Voice (text-to-speech) rate
Realtime speech model rateGPT-Live assistants (Beta)
Other phone rates
Other phone rates
Other phone rates
Other phone rates
Other phone rates
Phone number, per month
Add-on feeAny package
Add-on feeAny package; not available on legacy pricing
Add-on feeAny package
Volume discount

Billing rules

How a minute is billed
Billing ruleUsage only
Billing rule
Billing rule
Billing rule
Billing rule
Billing rulePro
Billing rule
Billing rule
Billing ruleGPT-Live assistants (Beta)

Capacity

Calls at once, includedNo Success Package (Usage only)
Calls at once, includedCore
Calls at once, includedPro
Calls at once, includedLegacy pricing
Each extra call at onceAny package, including Usage only
When you exceed the limit
Burst above the limit
Other limits
Other limitsCore
Other limits

Getting started

Free credit or trialNo Success Package (Usage only)
Sign up without salesNo Success Package (Usage only)
Sign up without salesCore, Pro

Compliance

HIPAA agreement (BAA)Any package, including Usage only
HIPAA agreement (BAA)
SOC 2All plans
Where data is processedPremier
Where data is processed
Other compliance statements
Other compliance statements
Other compliance statements
Other compliance statements
Other compliance statements

Product

Latency the vendor states
Latency the vendor states
Bring your own model, voice or carrier
Bring your own model, voice or carrier
White label for agencies
Uptime commitmentUsage only
Uptime commitmentPro
Uptime commitmentPremier
Support by planUsage only
Support by planCore
Support by planPro
Support by planPremier
Turn-taking setting
Turn-taking setting
Turn-taking setting
Turn-taking setting
Turn-taking setting
Turn-taking setting
Turn-taking setting
Turn-taking setting
Turn-taking setting
Turn-taking setting
How the vendor defines a resolved or contained call

Where Vapi's own pages say different things

Both statements are quoted with their pages. We do not pick one. Ask the vendor which applies before you rely on either.

What changed

  • 23 Sep 2026. Trust Center announced completed annual SOC 2 Type II and SOC 3 examinations covering July 1, 2025 to July 31, 2026. source
  • 15 Sep 2026. Trust Center announced a two-tier model separating Vapi Subprocessors from customer-selected Optional Providers. source
  • 14 Sep 2026. Terms of Service last updated. source

Pages we read

Lexicost has no commercial relationship with any vendor named on this site. Spotted an error? See corrections.

Source for every vendor figure on this page (99)
  1. Vapi, platform fee per minute: $0.05 per minute hosting fee (No Success Package (Usage only). pre-paid credits. Docs state usage-based billing stays active on Core, Pro and Premier).Pre-paid, $0.05/min Vapi hosting. vapi.ai/pricing Read 14 Nov 2026.
  2. Vapi, what that fee includes: None: speech, model, voice, transport extra (All packages).Four components: speech-to-text, the language model, text-to-speech, and transport. Each is billed at the provider's listed price with no markup. Vapi's $0.05/min hosting fee is added on top. vapi.ai/pricing Read 14 Nov 2026.
  3. Vapi, monthly plan fee: Core: $29 per month (Core. monthly, pre-paid, auto-renew. Added on top of usage-based billing).Core For builders & early teams $29 / month vapi.ai/pricing Read 14 Nov 2026.
  4. Vapi, minimum commitment: Pro plan only: $999 per month minimum (Pro. monthly. Applies only to the Pro Success Package, priced at 10% of the Vapi hosting fee with this monthly minimum; no minimum is printed on the No Success Package (usage only) or Core cards).10% of Vapi hosting fee $999/mo minimum vapi.ai/pricing Read 14 Nov 2026.
  5. Vapi, plan: Usage only: per-minute hosting fee plus model, voice, transcriber and telephony costs (Usage only (No Success Package)).With Usage only, you pay a per-minute hosting fee, plus the model, voice, transcriber, and telephony costs for each call. docs.vapi.ai/billing/pricing-and-success-packages Read 14 Nov 2026.
  6. Vapi, plan: Core: $29 per month, for builders and early teams (Core. monthly).Core For builders & early teams $29 / month vapi.ai/pricing Read 14 Nov 2026.
  7. Vapi, plan: Pro: 10% of the Vapi hosting fee, $999 per month minimum (Pro. monthly).10% of Vapi hosting fee $999/mo minimum vapi.ai/pricing Read 14 Nov 2026.
  8. Vapi, plan: Premier: custom pricing and terms through sales (Premier).Premier has custom pricing and terms, set with our sales team. docs.vapi.ai/billing/manage-packages-and-add-ons Read 14 Nov 2026.
  9. Vapi, how the AI model is charged: Passed through at cost (All packages).Models pass-through at cost vapi.ai/pricing Read 14 Nov 2026.
  10. Vapi, how the AI model is charged: Priced per million input and output tokens.Model 1 million input and output tokens docs.vapi.ai/assistants/model-intelligence/understanding-cost Read 14 Nov 2026.
  11. Vapi, how the AI model is charged: Estimates assume 150 output tokens a minute (Assumption used for the Model Intelligence cost estimate, not a billing rule).Vapi estimates 150 output tokens per minute based on typical generated speech in production call data. It prices those tokens at the model’s output rate. docs.vapi.ai/assistants/model-intelligence/understanding-cost Read 14 Nov 2026.
  12. Vapi, how the AI model is charged: Estimates assume a 50% cache hit rate (Assumption used for the Model Intelligence cost estimate, not a billing rule; only where the model supports prompt caching).When a model supports caching, Vapi blends its standard and cached rates based on an assumed 50% cache hit rate. Otherwise, Vapi uses the standard rate. docs.vapi.ai/assistants/model-intelligence/understanding-cost Read 14 Nov 2026.
  13. Vapi, how the AI model is charged: Estimates assume 5 model requests per turn (Assumption used for the Model Intelligence cost estimate, not a billing rule).Vapi estimates 5 requests per turn. docs.vapi.ai/assistants/model-intelligence/understanding-cost Read 14 Nov 2026.
  14. Vapi, aI model rate per minute: OpenAI: $0.0077 to $0.0452 per minute (calculator estimate) (Pricing-page calculator, Intelligence Model dropdown set to OpenAI (the default); pass-through cost; the page does not say which models the range covers).OpenAI $0.0077 - $0.0452/min vapi.ai/pricing Read 14 Nov 2026.
  15. Vapi, aI model rate per minute: Anthropic: $0.0082 to $0.0366 per minute (calculator estimate) (Pricing-page calculator, Intelligence Model dropdown set to Anthropic; pass-through cost; the page does not say which models the range covers).Anthropic $0.0082 - $0.0366/min vapi.ai/pricing Read 14 Nov 2026.
  16. Vapi, aI model rate per minute: Google: $0.006 to $0.0333 per minute (calculator estimate) (Pricing-page calculator, Intelligence Model dropdown set to Google; pass-through cost; the page does not say which models the range covers).Google $0.006 - $0.0333/min vapi.ai/pricing Read 14 Nov 2026.
  17. Vapi, speech-to-text rate: Deepgram: $0.0095 to $0.0099 per minute (calculator estimate) (Pricing-page calculator, Transcriber dropdown set to Deepgram (the default); pass-through cost; the page does not say which models the range covers).Deepgram $0.0095 - $0.0099/min vapi.ai/pricing Read 14 Nov 2026.
  18. Vapi, speech-to-text rate: AssemblyAI: $0.0052 to $0.0075 per minute (calculator estimate) (Pricing-page calculator, Transcriber dropdown set to AssemblyAI; pass-through cost; the page does not say which models the range covers).AssemblyAI $0.0052 - $0.0075/min vapi.ai/pricing Read 14 Nov 2026.
  19. Vapi, speech-to-text rate: OpenAI: $0.0118 to $0.0122 per minute (calculator estimate) (Pricing-page calculator, Transcriber dropdown set to OpenAI; pass-through cost; the page does not say which models the range covers).OpenAI $0.0118 - $0.0122/min vapi.ai/pricing Read 14 Nov 2026.
  20. Vapi, voice (text-to-speech) rate: ElevenLabs: $0.0146 to $0.0238 per minute (calculator estimate) (Pricing-page calculator, Voice dropdown set to ElevenLabs (the default); pass-through cost; the page does not say which voices the range covers).ElevenLabs $0.0146 - $0.0238/min vapi.ai/pricing Read 14 Nov 2026.
  21. Vapi, voice (text-to-speech) rate: Vapi voices: $0.007 to $0.0193 per minute (calculator estimate) (Pricing-page calculator, Voice dropdown set to Vapi; the page does not say which voices the range covers).Vapi $0.007 - $0.0193/min vapi.ai/pricing Read 14 Nov 2026.
  22. Vapi, voice (text-to-speech) rate: Cartesia: $0.0069 to $0.0129 per minute (calculator estimate) (Pricing-page calculator, Voice dropdown set to Cartesia; pass-through cost; the page does not say which voices the range covers).Cartesia $0.0069 - $0.0129/min vapi.ai/pricing Read 14 Nov 2026.
  23. Vapi, realtime speech model rate: GPT-Live voice: $0.05 per minute (worked example) (GPT-Live assistants (Beta). Illustrative estimate using public rates as of September 30, 2026).GPT-Live voice 4 minutes at $0.05 per minute $0.20 docs.vapi.ai/gpt-live/testing Read 14 Nov 2026.
  24. Vapi, uS phone minutes: Free on Vapi numbers (Vapi Telephony / SIP) (Vapi Telephony / SIP (free Vapi numbers). Transport Fee in the pricing-page estimate; 1 included Vapi phone number without a Success Package, 5 on Core, 10 on Pro; calls still pay the hosting fee and model costs).Transport Fee Vapi Telephony / SIP $0 vapi.ai/pricing Read 14 Nov 2026.
  25. Vapi, other phone rates: Twilio inbound: $0.008 per minute, charged by Twilio (Charged by provider, not Vapi).Twilio Inbound $0.008 / min vapi.ai/pricing Read 14 Nov 2026.
  26. Vapi, other phone rates: Twilio outbound: $0.014 per minute, charged by Twilio (Charged by provider, not Vapi).Twilio Outbound $0.014 / min vapi.ai/pricing Read 14 Nov 2026.
  27. Vapi, other phone rates: Vonage: $0.00814 per minute, charged by Vonage (Charged by provider, not Vapi).Vonage $0.00814 / min vapi.ai/pricing Read 14 Nov 2026.
  28. Vapi, other phone rates: Telnyx: $0.0055 per minute, charged by Telnyx (Charged by provider, not Vapi).Telnyx $0.0055 / min vapi.ai/pricing Read 14 Nov 2026.
  29. Vapi, other phone rates: Vapi WebSockets: free.Vapi WebSockets Free vapi.ai/pricing Read 14 Nov 2026.
  30. Vapi, how a minute is billed: Not publicly listed. Read 14 Nov 2026.
  31. Vapi, calls at once, included: 4 calls at once (No Success Package (Usage only). Per subscription, shared by all its organizations).4 concurrent calls vapi.ai/pricing Read 14 Nov 2026.
  32. Vapi, calls at once, included: 10 calls at once (Core. Per subscription, shared by all its organizations).10 concurrent calls vapi.ai/pricing Read 14 Nov 2026.
  33. Vapi, calls at once, included: 30 calls at once (Pro. Per subscription, shared by all its organizations).30 concurrent calls vapi.ai/pricing Read 14 Nov 2026.
  34. Vapi, calls at once, included: Legacy accounts: 10 per organization (Legacy pricing. Each organization has its own subscription and limit).On legacy accounts, each organization has its own subscription, so each organization has its own limit of 10 included concurrent call lines. docs.vapi.ai/calls/call-concurrency Read 14 Nov 2026.
  35. Vapi, each extra call at once: $10 per extra call line per month (Any package, including Usage only. monthly, charged to credits on the first day of each month).Additional concurrency: $10 per concurrent call line per month. vapi.ai/pricing Read 14 Nov 2026.
  36. Vapi, when you exceed the limit: New outbound calls cannot start and inbound calls are not accepted until a slot frees.When all slots are busy, you cannot start an outbound call or accept an inbound call until a slot becomes available. docs.vapi.ai/calls/call-concurrency Read 14 Nov 2026.
  37. Vapi, burst above the limit: Not publicly listed. Read 14 Nov 2026.
  38. Vapi, other limits: Concurrency is counted per subscription and shared by its organizations (Current pricing model).Call concurrency is measured at the subscription level, and all organizations in a subscription share it. docs.vapi.ai/calls/call-concurrency Read 14 Nov 2026.
  39. Vapi, other limits: Core: 2 organizations (pricing page) (Core).Core For builders & early teams $29 / month 2 organizations Also: Organizations 2 5 10 Custom docs.vapi.ai/billing/pricing-and-success-packages vapi.ai/pricing Read 14 Nov 2026.
  40. Vapi, other limits: Included Vapi phone numbers: 1 (Usage only), 5 (Core), 10 (Pro), custom (Premier).Vapi phone numbers 1 5 10 Custom docs.vapi.ai/billing/pricing-and-success-packages Read 14 Nov 2026.
  41. Vapi, phone number, per month: Not publicly listed. Read 14 Nov 2026.
  42. Vapi, free credit or trial: $5 free credits (No Success Package (Usage only)).$5 free credits to get you started vapi.ai/pricing Read 14 Nov 2026.
  43. Vapi, sign up without sales: Yes: start free and pay per minute (No Success Package (Usage only)).Get started for free, and then pay only for the minutes that run vapi.ai/pricing Read 14 Nov 2026.
  44. Vapi, sign up without sales: Core and Pro can be bought in the Dashboard (Core, Pro).Buy Core or Pro directly from the Dashboard when you need more support, capacity, or service commitments. docs.vapi.ai/billing/manage-packages-and-add-ons Read 14 Nov 2026.
  45. Vapi, hIPAA agreement (BAA): Yes, as a $2,000 per month add-on (Any package, including Usage only. monthly. Per organization).HIPAA-eligible data handling is a $2,000/month add-on per organization and comes with a standard DPA and BAA. vapi.ai/pricing Read 14 Nov 2026.
  46. Vapi, hIPAA agreement (BAA): BAA emailed on enabling HIPAA in the Dashboard.To enable HIPAA, select Enable HIPAA and enter the signatory name and legal entity name. Vapi emails a Business Associate Agreement (BAA) to your billing email address. docs.vapi.ai/billing/manage-packages-and-add-ons Read 14 Nov 2026.
  47. Vapi, add-on fee: HIPAA compliance: $2,000 per month (Any package).HIPAA compliance HIPAA eligible data handling. Standard Data Processing Agreement and Business Associate Agreement $2,000 / month vapi.ai/pricing Read 14 Nov 2026.
  48. Vapi, add-on fee: Extra organization: $20 per organization per month (Any package; not available on legacy pricing).Additional organizations: extra workspaces beyond your package default, $20 per organization per month. vapi.ai/pricing Read 14 Nov 2026.
  49. Vapi, add-on fee: 60-day data retention add-on (price not shown) (Any package).HIPAA and 60-day data retention are available with any package. Zero data retention (ZDR) is included in Core and Pro, and is not available as an add-on on Usage only. docs.vapi.ai/billing/manage-packages-and-add-ons Read 14 Nov 2026.
  50. Vapi, volume discount: Volume pricing available as you scale (no figures published).Yes, volume pricing is available as you scale. See the Vapi pricing page for details. docs.vapi.ai/faq Read 14 Nov 2026.
  51. Vapi, billing rule: Every minute an assistant runs is billed (Usage only).Every minute your assistants run is billed. docs.vapi.ai/billing/pricing-and-success-packages Read 14 Nov 2026.
  52. Vapi, billing rule: Incoming calls are not stopped unless you set limits; you pay for all minutes.Unless you have limited the traffic flow to certain limits, we do not stop incoming voice calls to our Platform. As such, you are responsible for payment of all the minutes utilized on our Platform, regardless of whether the voice traffic exceeded your contemplation. vapi.ai/terms-of-service Read 14 Nov 2026.
  53. Vapi, billing rule: Per-minute and metered charges are non-refundable.This paragraph does not apply to fees calculated on a consumption or usage basis (including per-minute or metered charges under Section 10(d)), which are earned as incurred and are non-refundable except as required by law. vapi.ai/terms-of-service Read 14 Nov 2026.
  54. Vapi, billing rule: At a $0 balance with auto-reload off, the subscription is frozen and new calls cannot start (pre-paid credits).If the subscription’s credit balance reaches $0 and auto-reload is off, Vapi freezes the subscription. You cannot start new calls or other billable usage until you purchase more credits. docs.vapi.ai/billing/manage-billing-and-credits Read 14 Nov 2026.
  55. Vapi, billing rule: Add-ons are charged to credits on the first day of each month.Add-ons are charged to your Vapi credits on the first day of each month. docs.vapi.ai/billing/manage-billing-and-credits Read 14 Nov 2026.
  56. Vapi, billing rule: Pro fee is a percentage of hosting fees only, not model costs (Pro).It's a percentage of your Vapi hosting fees, model costs are not included. vapi.ai/pricing Read 14 Nov 2026.
  57. Vapi, billing rule: Transcription is billed per audio minute and covers both caller and assistant audio.Transcription is billed per minute of audio. Vapi transcribes audio from both the caller and the assistant. The estimate therefore includes two audio channels. docs.vapi.ai/assistants/model-intelligence/understanding-cost Read 14 Nov 2026.
  58. Vapi, billing rule: Advance notice and an option to cancel for material pricing changes.If we introduce material changes to pricing, we will notify you in advance and will provide you an option to cancel the Services. vapi.ai/terms-of-service Read 14 Nov 2026.
  59. Vapi, billing rule: GPT-Live voice time is billed by the second, including silence and waiting (GPT-Live assistants (Beta)).A GPT-Live call is billed for voice time by the second, including silence and waiting, plus Vapi’s platform fee, the reasoner’s token usage, and telephony. docs.vapi.ai/gpt-live/overview Read 14 Nov 2026.
  60. Vapi, sOC 2: SOC 2 Type II attested (All plans).Vapi, as a Voice Agent Platform, is a SOC 2 Type II-attested service organization and undergoes annual third-party audits to validate its compliance with the framework. docs.vapi.ai/security-and-privacy/soc Read 14 Nov 2026.
  61. Vapi, where data is processed: EU region only with Premier (Premier).Vapi’s EU region is available only with the Premier Success Package. Contact sales to discuss EU access. docs.vapi.ai/security-and-privacy/eu-region Read 14 Nov 2026.
  62. Vapi, where data is processed: No on-premise deployment.No, Vapi does not support on-premise deployments. docs.vapi.ai/faq Read 14 Nov 2026.
  63. Vapi, other compliance statements: States SOC 2, HIPAA and PCI compliance.SOC 2, HIPAA, and PCI compliant vapi.ai Read 14 Nov 2026.
  64. Vapi, other compliance statements: Card data only with a separate written agreement.You may not submit, transmit, or store payment card information ("PCI Data") through the Services unless you have separately agreed in writing with Vapi on the applicable scope, controls, and obligations. vapi.ai/terms-of-service Read 14 Nov 2026.
  65. Vapi, other compliance statements: States GDPR compliance.GDPR compliant for handling EU data docs.vapi.ai/faq Read 14 Nov 2026.
  66. Vapi, other compliance statements: HIPAA mode and Zero Data Retention cannot be on together.HIPAA mode and Zero Data Retention (ZDR) are mutually exclusive. Disable one before enabling the other. docs.vapi.ai/security-and-privacy/hipaa Read 14 Nov 2026.
  67. Vapi, other compliance statements: Raw data retention: 14 days (Usage only), 30 days (Core), 180 days (Pro), custom (Premier).Raw data retention 14 days 30 days 180 days Custom docs.vapi.ai/billing/pricing-and-success-packages Read 14 Nov 2026.
  68. Vapi, latency the vendor states: Sub-500 ms latency; what it measures is not stated (Homepage claim; no measure or conditions given).Scale up to millions of calls with sub-500ms latency. vapi.ai Read 14 Nov 2026.
  69. Vapi, latency the vendor states: Typically around 800 ms for end-to-end voice processing (End-to-end voice processing, typical figure; no test conditions given).Yes, Vapi is designed to achieve low latency, typically around 800 milliseconds for end-to-end voice processing. docs.vapi.ai/faq Read 14 Nov 2026.
  70. Vapi, bring your own model, voice or carrier: Bring your own provider keys; that provider then bills you directly.Once your API key is validated, you won’t be charged when using that provider through Vapi. Instead, you’ll be charged directly by the provider. docs.vapi.ai/customization/provider-keys Read 14 Nov 2026.
  71. Vapi, bring your own model, voice or carrier: Import numbers from Twilio, Telnyx or DIDWW.Free Vapi phone numbers use US area codes, and only inbound calls are supported. To make outbound calls, import a number from Twilio, Telnyx, or DIDWW. docs.vapi.ai/phone-calling Read 14 Nov 2026.
  72. Vapi, white label for agencies: Yes, through the API without Vapi branding.Yes, Vapi is a API-first platform. This means can bake it in your product without using Vapi branding. docs.vapi.ai/faq Read 14 Nov 2026.
  73. Vapi, uptime commitment: No uptime commitment on Usage only (Usage only).Usage only does not include a response or uptime service-level agreement. docs.vapi.ai/billing/pricing-and-success-packages Read 14 Nov 2026.
  74. Vapi, uptime commitment: Pro: 99% uptime SLA (Pro).99% uptime SLA vapi.ai/pricing Read 14 Nov 2026.
  75. Vapi, uptime commitment: Premier: 99.9% uptime SLA (Premier).99.9% uptime SLA vapi.ai/pricing Read 14 Nov 2026.
  76. Vapi, support by plan: Usage only: docs, Discord community and email, no response SLA (Usage only).Usage only Documentation, peer-to-peer Discord community, and email support No response or uptime service-level agreement Included with every subscription docs.vapi.ai/billing/pricing-and-success-packages Read 14 Nov 2026.
  77. Vapi, support by plan: Core: two-business-day response SLA (Core).Core Everything in Usage Two-business-day response service-level agreement Buy in the Dashboard docs.vapi.ai/billing/pricing-and-success-packages Read 14 Nov 2026.
  78. Vapi, support by plan: Pro: dedicated Slack channel, one-business-day response SLA (Pro).Pro Everything in Core, plus a dedicated Slack channel One-business-day response and 99% uptime service-level agreements Buy in the Dashboard docs.vapi.ai/billing/pricing-and-success-packages Read 14 Nov 2026.
  79. Vapi, support by plan: Premier: named support team, one-hour P0 response SLA (Premier).Premier Everything in Pro, plus a named support team One-hour P0 response and 99.9% uptime service-level agreements Contact Sales docs.vapi.ai/billing/pricing-and-success-packages Read 14 Nov 2026.
  80. Vapi, where the vendor's own pages disagree: The pricing page lists 2 organizations for Core, while the billing docs list 5. (Core).Core For builders & early teams $29 / month 2 organizations Also: Organizations 2 5 10 Custom docs.vapi.ai/billing/pricing-and-success-packages vapi.ai/pricing Read 14 Nov 2026.
  81. Vapi, turn-taking setting: waitSeconds: how long the assistant waits before speaking; default 0.4 seconds (0 to 5 seconds; part of startSpeakingPlan).This is how long assistant waits before speaking. Defaults to 0.4. docs.vapi.ai/api-reference/assistants/create Read 14 Nov 2026.
  82. Vapi, turn-taking setting: Smart endpointing: Off by default; uses the transcriber's end-of-turn detection, else LiveKit (English) or Vapi (Part of startSpeakingPlan).Off: Disabled by default. When smart endpointing is set to “Off”, the system will automatically use the transcriber’s end-of-turn detection if available. If no transcriber EOT detection is available, the system defaults to LiveKit if the language is set to English or to Vapi’s standard endpointing mode. docs.vapi.ai/customization/speech-configuration Read 14 Nov 2026.
  83. Vapi, turn-taking setting: waitFunction (LiveKit): how long the bot waits to start speaking, from how likely the user has finished (Only with LiveKit smart endpointing; maps a probability from 0 to 1 to a wait in milliseconds).LiveKit Smart Endpointing Configuration: When using LiveKit, you can customize the waitFunction parameter which determines how long the bot will wait to start speaking based on the likelihood that the user has finished speaking: docs.vapi.ai/customization/speech-configuration Read 14 Nov 2026.
  84. Vapi, turn-taking setting: waitFunction default: the Dashboard starts on the Aggressive preset (speech configuration guide) (LiveKit smart endpointing only).The Dashboard currently starts with the Aggressive preset when you select LiveKit; choose a preset or enter your own expression in advanced mode. Also: Under the hood, this is parsed into a mathjs expression. Whatever you use to write your expression needs to be valid with respect to mathjs @default “20 + 500 * sqrt(x) + 2500 * x^3” docs.vapi.ai/api-reference/assistants/create Also: Aggressive (Fast Response): "waitFunction": "2000 / (1 + exp(-10 * (x - 0.5)))" docs.vapi.ai/customization/voice-pipeline-configuration docs.vapi.ai/customization/speech-configuration Read 14 Nov 2026.
  85. Vapi, turn-taking setting: customEndpointingRules: timeouts set by a regex on the customer's speech or the assistant's last message (Part of startSpeakingPlan; list of rules, each with its own timeout in seconds).These are the custom endpointing rules to set an endpointing timeout based on a regex on the customer’s speech or the assistant’s last message. docs.vapi.ai/api-reference/assistants/create Read 14 Nov 2026.
  86. Vapi, turn-taking setting: onPunctuationSeconds: minimum wait after transcription ending with punctuation, before the model; default 0.1 s (0 to 3 seconds; part of startSpeakingPlan.transcriptionEndpointingPlan, which the API reference says is used only when smartEndpointingPlan is not set and the transcriber has no built-in endpointing).The minimum number of seconds to wait after transcription ending with punctuation before sending a request to the model. Defaults to 0.1. docs.vapi.ai/api-reference/assistants/create Read 14 Nov 2026.
  87. Vapi, turn-taking setting: onNoPunctuationSeconds: minimum wait after transcription ending without punctuation; default 1.5 s (0 to 3 seconds; part of startSpeakingPlan.transcriptionEndpointingPlan).The minimum number of seconds to wait after transcription ending without punctuation before sending a request to the model. Defaults to 1.5. docs.vapi.ai/api-reference/assistants/create Read 14 Nov 2026.
  88. Vapi, turn-taking setting: onNumberSeconds: the wait after a transcript that ends with a number; default 0.5 seconds (voice pipeline guide) (0 to 3 seconds; part of startSpeakingPlan.transcriptionEndpointingPlan).Number detection: If the latest message ends with a number, waits for onNumberSeconds (default: 0.5) Also: The minimum number of seconds to wait after transcription ending with a number before sending a request to the model. Defaults to 0.4. This setting exists because the transcriber will sometimes punctuate the transcription ending with a number, even though the customer hasn’t uttered the full number. This happens commonly for long numbers when the customer reads the number in chunks. @default 0.5 docs.vapi.ai/api-reference/assistants/create Also: The system detects the number “123456789” and waits for 0.5 seconds (WaitSeconds) to ensure the customer isn’t still speaking. docs.vapi.ai/customization/speech-configuration docs.vapi.ai/customization/voice-pipeline-configuration Read 14 Nov 2026.
  89. Vapi, where the vendor's own pages disagree: The API reference text gives onNumberSeconds a default of 0.4 seconds, while its own @default tag and the voice pipeline guide give 0.5, and the speech configuration guide calls the 0.5-second wait after a number WaitSeconds..Number detection: If the latest message ends with a number, waits for onNumberSeconds (default: 0.5) Also: The minimum number of seconds to wait after transcription ending with a number before sending a request to the model. Defaults to 0.4. This setting exists because the transcriber will sometimes punctuate the transcription ending with a number, even though the customer hasn’t uttered the full number. This happens commonly for long numbers when the customer reads the number in chunks. @default 0.5 docs.vapi.ai/api-reference/assistants/create Also: The system detects the number “123456789” and waits for 0.5 seconds (WaitSeconds) to ensure the customer isn’t still speaking. docs.vapi.ai/customization/speech-configuration docs.vapi.ai/customization/voice-pipeline-configuration Read 14 Nov 2026.
  90. Vapi, where the vendor's own pages disagree: The speech configuration guide says the Dashboard starts LiveKit’s waitFunction on the Aggressive preset, which the voice pipeline guide gives as 2000 / (1 + exp(-10 * (x - 0.5))), while the API reference gives the default as 20 + 500 * sqrt(x) + 2500 * x^3..The Dashboard currently starts with the Aggressive preset when you select LiveKit; choose a preset or enter your own expression in advanced mode. Also: Under the hood, this is parsed into a mathjs expression. Whatever you use to write your expression needs to be valid with respect to mathjs @default “20 + 500 * sqrt(x) + 2500 * x^3” docs.vapi.ai/api-reference/assistants/create Also: Aggressive (Fast Response): "waitFunction": "2000 / (1 + exp(-10 * (x - 0.5)))" docs.vapi.ai/customization/voice-pipeline-configuration docs.vapi.ai/customization/speech-configuration Read 14 Nov 2026.
  91. Vapi, turn-taking setting: numWords: how many words the customer must say before the assistant stops talking; default 0 (0 to 10 words; part of stopSpeakingPlan; at 0, voiceSeconds is also used).This is the number of words that the customer has to say before the assistant will stop talking. Words like “stop”, “actually”, “no”, etc. will always interrupt immediately regardless of this value. Words like “okay”, “yeah”, “right” will never interrupt. When set to 0, voiceSeconds is used in addition to the transcriptions to determine the customer has started speaking. Defaults to 0. docs.vapi.ai/api-reference/assistants/create Read 14 Nov 2026.
  92. Vapi, turn-taking setting: voiceSeconds: how long voice activity must be detected before an interruption, if numWords is 0; default 0.2 s (0 to 0.5 seconds; part of stopSpeakingPlan; used only when numWords is 0).VAD duration threshold when numWords = 0. Determines how long voice activity must be detected before triggering an interruption. Range: 0-0.5 seconds (Default: 0.2) docs.vapi.ai/customization/voice-pipeline-configuration Read 14 Nov 2026.
  93. Vapi, turn-taking setting: backoffSeconds: how long the assistant waits before talking again after being interrupted; default 1 second (0 to 10 seconds; part of stopSpeakingPlan).This is the seconds to wait before the assistant will start talking again after being interrupted. Defaults to 1. docs.vapi.ai/api-reference/assistants/create Read 14 Nov 2026.
  94. Vapi, turn-taking setting: acknowledgementPhrases: phrases that never interrupt the assistant, such as "okay" and "uh-huh"; 22 by default (Part of stopSpeakingPlan; list of strings).Defaults to ["i understand","i see","i got it","i hear you","im listening","im with you","right","okay","ok","sure","alright","got it","understood","yeah","yes","uh-huh","mm-hmm","gotcha","mhmm","ah","yeah okay","yeah sure"] These are the phrases that will never interrupt the assistant, even if numWords threshold is met. These are typically acknowledgement or backchanneling phrases. docs.vapi.ai/api-reference/assistants/create Read 14 Nov 2026.
  95. Vapi, turn-taking setting: interruptionPhrases: phrases that always interrupt the assistant at once, such as "stop"; 20 by default (Part of stopSpeakingPlan; list of strings).Defaults to ["stop","shut","up","enough","quiet","silence","but","dont","not","no","hold","wait","cut","pause","nope","nah","nevermind","never","bad","actually"] These are the phrases that will always interrupt the assistant immediately, regardless of numWords. These are typically phrases indicating disagreement or desire to stop. docs.vapi.ai/api-reference/assistants/create Read 14 Nov 2026.
  96. Vapi, turn-taking setting: firstMessageMode: whether the assistant speaks first or waits for the user; default assistant-speaks-first (Allowed values: assistant-speaks-first, assistant-speaks-first-with-model-generated-message, assistant-waits-for-user).This is the mode for the first message. Default is ‘assistant-speaks-first’. Use: ‘assistant-speaks-first’ to have the assistant speak first. ‘assistant-waits-for-user’ to have the assistant wait for the user to speak first. docs.vapi.ai/api-reference/assistants/create Read 14 Nov 2026.
  97. Vapi, turn-taking setting: eotThreshold (Deepgram Flux): confidence level required to trigger end-of-turn detection; default 0.7 (Deepgram Flux models only; set on the transcriber, with no smartEndpointingPlan).eotThreshold (Default: 0.7): Confidence level required to trigger end-of-turn detection docs.vapi.ai/customization/voice-pipeline-configuration Read 14 Nov 2026.
  98. Vapi, turn-taking setting: eotTimeoutMs (Deepgram Flux): maximum wait in milliseconds before forcing the turn to end; default 5,000 (Deepgram Flux models only; milliseconds; set on the transcriber).eotTimeoutMs (Default: 5000): Maximum wait time in milliseconds before forcing turn end docs.vapi.ai/customization/voice-pipeline-configuration Read 14 Nov 2026.
  99. Vapi, how the vendor defines a resolved or contained call: Success evaluation: by default a model judges success against objectives inferred from the system prompt (analysisPlan success evaluation; the prompt and rubric can be customized, and the page recommends structured outputs or a scorecard as the equivalent).Used to determine if the call was successful, stored in call.analysis.successEvaluation. Default prompt: You are an expert call evaluator. You will be given a transcript of a call and the system prompt of the AI participant. Determine if the call was successful based on the objectives inferred from the system prompt. docs.vapi.ai/assistants/call-analysis Read 14 Nov 2026.