Vapi · Pricing · Checked Oct 27, 2026
Vapi AI pricing: the real cost per minute in 2026
Vapi AI charges $0.05/min in hosting. Transcriber, LLM, voice and phone transport are passed through at cost on top. Vapi's own calculator puts 1,000 minutes at $82–$129/month. With 10 calls at once the same month runs $110.80–$157.90. Checked Oct 27, 2026 on vapi.ai/pricing.
On our stack
$110.80–$157.90
1,000 min, computed
Included concurrency
Vapi pricing at a glance
| Item | Value |
|---|---|
| Platform fee per minute (hosting) | $0.05/min |
| Vapi's own calculator, 1,000 min | $82–$129/month |
| Monthly fee, no package | $0/month |
| Free credit | $5 credits |
| Concurrency included | 4 / 10 / 30 |
| Extra concurrency | $10/line/month |
| Phone transport | Bring your own carrier No Vapi fee on your own SIP; carriers bill you: Twilio $0.008/min in, $0.014/min out; Vonage $0.00814/min; Telnyx $0.0055/min |
Sources: vapi.ai · Oct 27, 2026
What a minute on Vapi actually costs
Cost per minute = hosting + transcriber + LLM tokens + voice + transport. Then add concurrency lines or a package.
| Component | Per minute (USD) |
|---|---|
| Hosting | $0.05 |
| Transcriber (Deepgram) | $0.0095 to $0.0099 |
| LLM (OpenAI) | $0.0077 to $0.0452 |
| Voice (ElevenLabs) | $0.0146 to $0.0238 |
| Phone transport | Bring your own carrier No Vapi fee on your own SIP; carriers bill you: Twilio $0.008/min in, $0.014/min out; Vonage $0.00814/min; Telnyx $0.0055/min |
Sources: vapi.ai · Oct 27, 2026
Hosting fee
$0.05/min, prepaid. This is the only part Vapi itself charges per minute.
Transcriber
Deepgram through Vapi is $0.0095–$0.0099/min, passed through.
LLM tokens
Vapi bills the LLM by tokens. Its docs estimate cost with 150 output tokens/min; 5 requests per turn; 50% cache-hit assumption; input tokens depend on your prompt. The calculator shows $0.0077–$0.0452/min for OpenAI models.
Voice
ElevenLabs through Vapi is $0.0146–$0.0238/min.
Phone transport
No Vapi fee on your own SIP; carriers bill you: Twilio $0.008/min in, $0.014/min out; Vonage $0.00814/min; Telnyx $0.0055/min. Twilio's own US rates, for reference, are $0.0085/min inbound and $0.014/min outbound local, plus $1.15/month per number.
Vapi plans: usage only, Core, Pro, Premier
| Plan | Monthly fee | What else |
|---|---|---|
| No package | $0 /mo | 4 concurrent calls, 1 number, 2 orgs, 14-day retention |
| Core | $29 /mo | 10 concurrent calls, 5 numbers, 30-day retention, ZDR, email support 2-day SLA |
| Pro | 10% of hosting fee, $999 minimum | 30 concurrent calls, 10 numbers, 180-day retention, Slack support, 99% SLA |
| Premier | Not published Vapi: Custom, direct sales only | Direct sales only |
Sources: vapi.ai · Oct 27, 2026
Vapi concurrency limit and extra lines
Included concurrency is 4 / 10 / 30. Extra lines are $10/line/month. For 10 calls at once, six extra lines cost more than Core at $29/month, so the volume tables use Core. Premier is Not published by Vapi. Custom, direct sales only.
Vapi cost at 500 to 10,000 minutes a month
Scenario (our assumptions, not vendor figures): inbound calls on one US local number; average call 3 minutes; the business needs 10 calls at once; a GPT-4.1-mini-class LLM; each platform's cheapest published voice. Where a platform bills the LLM by token and publishes no per-minute figure, we use Retell's published GPT 4.1 mini rate ($0.0128/min) as a stand-in and label it assumption. For Vapi we use the low and high ends of Vapi’s own calculator instead of the Retell stand-in.
| Line item | 500 min | 1k min | 5k min | 10k min |
|---|---|---|---|---|
| Hosting | $25.00 | $50.00 | $250.00 | $500.00 |
| STT (Deepgram) | $4.75 | $9.50 | $47.50 | $95.00 |
| LLM (OpenAI, Vapi calculator range) | $3.85 | $7.70 | $38.50 | $77.00 |
| TTS (ElevenLabs, Vapi calculator range) | $7.30 | $14.60 | $73.00 | $146.00 |
| Telephony (Vapi numbers, $0) | $0.00 | $0.00 | $0.00 | $0.00 |
| Concurrency for 10 calls: Core plan $29 10 concurrent | $29.00 | $29.00 | $29.00 | $29.00 |
| Monthly total | $69.90 | $110.80 | $438.00 | $847.00 |
| Effective per minute | $0.1398 | $0.1108 | $0.0876 | $0.0847 |
1,000 min. 5,000 min. 10,000 min Sources: vapi.ai · Oct 27, 2026
| Line item | 500 min | 1k min | 5k min | 10k min |
|---|---|---|---|---|
| Hosting | $25.00 | $50.00 | $250.00 | $500.00 |
| STT (Deepgram) | $4.95 | $9.90 | $49.50 | $99.00 |
| LLM (OpenAI, Vapi calculator range) | $22.60 | $45.20 | $226.00 | $452.00 |
| TTS (ElevenLabs, Vapi calculator range) | $11.90 | $23.80 | $119.00 | $238.00 |
| Telephony (Vapi numbers, $0) | $0.00 | $0.00 | $0.00 | $0.00 |
| Concurrency for 10 calls: Core plan $29 10 concurrent | $29.00 | $29.00 | $29.00 | $29.00 |
| Monthly total | $93.45 | $157.90 | $673.50 | $1,318.00 |
| Effective per minute | $0.1869 | $0.1579 | $0.1347 | $0.1318 |
1,000 min. 5,000 min. 10,000 min Sources: vapi.ai · Oct 27, 2026
Vapi’s own 1,000-minute figure excludes concurrency. Ours adds Core. Run your own numbers.
Hidden costs: HIPAA, organizations, prepaid credits
HIPAA is $2,000/month. Extra organizations are $20/org/month. Usage is paid from prepaid credits. Pro is 10% of hosting fee, $999 minimum, not a flat seat price.
Our Vapi test calls
We have not run billed test calls on Vapi. This phase records list prices only: no sign-ups and no spend. Until a billed call is logged, the tables above are the sourced model, not a measured invoice.
Vapi price changes
Oct 27, 2026: first recorded on this site, from vapi.ai/pricing.
Is Vapi worth it?
Vapi suits teams who want to pick every provider and can live with a bill that moves with token use. Above the free concurrency, the package fee is part of the real price. On our model Vapi costs more than Retell at 500 and 1,000 minutes, and its low end is cheaper at 5,000 and up. For one bundled rate, see Bland pricing. See also Retell pricing, Vapi vs Retell AI and Vapi alternatives.
FAQ
Is Vapi free?
No. New accounts get $5 credits. After that you pay $0.05/min hosting plus provider costs. Source: https://vapi.ai/pricing (checked Oct 27, 2026).
How much does Vapi cost per minute?
$0.05/min hosting plus pass-through providers. Vapi's own calculator puts 1,000 minutes at $82–$129/month before concurrency. Our 10-call stack is $110.80–$157.90. Source: https://vapi.ai/pricing (checked Oct 27, 2026).
What is Vapi's concurrency limit?
Included concurrency is 4 / 10 / 30 (no package / Core / Pro). Extra lines are $10/line/month. Source: https://vapi.ai/pricing (checked Oct 27, 2026).
Sources
- vapi.ai ·
- docs.vapi.ai ·
- twilio.com ·