Retell AI Review 2026: Pricing, Features and Verdict
The best all round voice agent platform on published facts, once you price the parts the headline leaves out.
By , research by the BestBots team · Updated Oct 18, 2026
Fact checked against each company's published pricing, documentation and verified user ratings. How we score
Retell AI is our top pick for AI voice agents in October 2026. It publishes every cost layer on one page, includes 20 concurrent calls and HIPAA at no extra charge, and lets non developers build in a visual canvas. The catch is the headline: $0.055 per minute is only the voice engine, and a real stack runs $0.09 to $0.17.
The most complete self serve voice agent platform on published facts: itemized per minute pricing, 20 free concurrent calls, SOC 2 Type II and HIPAA without an add on fee. The tradeoff is that the $0.055 headline rate is only the voice engine, so a real stack lands near $0.09 to $0.17 per minute once you add an LLM, a voice and telephony.
No user reviews yet. Write a review
- Starting price
- $0.055 per minute
- Pricing model
- per minute, pay as you go
- Free plan
- No
- Free trial
- $10 in free credits
- Per minute cost
- $0.055 voice infra plus LLM ($0.0016 to $0.32), voice ($0.015 to $0.10) and telephony ($0.015 US Twilio or Telnyx)
- Latency
- About 600 ms, per Retell's homepage
- Telephony
- Retell numbers via Twilio, Vonage, or your own SIP trunk at no Retell charge
- Concurrency
- 20 concurrent calls free, then $8 per line per month
- Languages
- Multilingual agents supported; the dashboard lists current languages
- HIPAA
- Yes, listed as HIPAA ready
- SOC 2
- SOC 2 Type II
What it is
Retell AI is a platform for building, testing, deploying and monitoring AI phone agents (and chat agents). It handles the real time parts of a call: listening, turn taking, interruptions, sending text to an LLM, and speaking the reply in a synthetic voice. You define what the agent says and does, connect it to tools like a calendar or CRM, and attach a phone number.
There are two ways to build. The drag and drop canvas lets you lay out a call flow with branches, transfers and tool calls. Prompt based agents let you describe the job in plain language. Developers can do everything through the API. Retell's docs walk through a first phone agent in about 15 minutes: create an account, pick a template, test in the dashboard and deploy to a number.
Pricing explained
Retell is pay as you go with no monthly fee, plus custom Enterprise pricing. New accounts get $10 in credits. Here is every line item listed on its pricing page in October 2026.
| Component | Price |
|---|---|
| Voice infra (includes speech to text) | $0.055 per minute |
| LLM, Claude 5.5 Haiku | $0.0032 per minute |
| LLM, Gemini 3.5 Flash Lite | $0.0096 per minute |
| LLM, GPT 5.6 Terra or Claude 5 Sonnet | $0.064 per minute |
| LLM, GPT Realtime mini | $0.07 per minute |
| LLM, GPT 5.4 | $0.080 per minute |
| LLM, GPT 5.5 | $0.16 per minute |
| Voice, Retell or OpenAI | $0.015 per minute |
| Voice, ElevenLabs Flash | $0.040 per minute |
| Voice, ElevenLabs v3 | $0.10 per minute |
| Telephony, US Twilio | $0.015 per minute |
| SIP trunking or custom telephony | No charge |
| Concurrency | 20 free, then $8 per line per month |
| Knowledge base | $8 per month each |
| Batch calls | $0.005 per dial |
| Verified phone number | $10 per month |
Three realistic stacks. Cheapest: Claude 5.5 Haiku, Retell voice, US Twilio = $0.055 + $0.0032 + $0.015 + $0.015 = $0.0882 per minute. Budget with a stronger small model: Gemini 3.5 Flash Lite instead = $0.0946. Premium: Claude 5 Sonnet with ElevenLabs Flash = $0.174. If you bring your own SIP trunk, subtract $0.015 and pay your carrier instead.
Monthly bills. At 5,000 minutes: $441 cheapest, $473 budget, $870 premium. At 20,000 minutes: $1,764, $1,892 and $3,480. Add $8 per month for each knowledge base and $8 per line above 20 concurrent calls. Retell's homepage summarizes the range as $0.07 to $0.31 per minute.
What surprises people. The LLM and voice can cost more than the platform. GPT 5.5 alone ($0.16) is nearly three times the voice infra rate. Pick the smallest model that handles your script and test it before reaching for a frontier model.
These rates feed the AI Price Index, and the cost per minute breakdown compares them with every other platform.
Features
- Builder. Drag and drop canvas for flows, prompt agents for simpler jobs, plus testing, deployment and QA tools in the dashboard.
- Latency. Retell states about 600 ms.
- Telephony. Retell numbers via Twilio, Vonage or any SIP trunk, with no Retell fee for custom telephony.
- Concurrency. 20 concurrent calls included, which covers most small and mid sized inbound lines without an upgrade.
- Languages. Multilingual agents are supported, and the language picker in the dashboard is the source of truth; Retell does not publish a count.
- Compliance. SOC 2 Type II, HIPAA, GDPR and ISO 27001, plus SSO, role based access and personal information redaction. On premises deployment is available on Enterprise.
- Integrations. HubSpot, Cal.com, Make, n8n, GoHighLevel and custom APIs as tools.
- Benchmarks. Retell cites an independent Cekura benchmark showing 95.7% workflow accuracy across 414 calls, highest of six platforms. That is a vendor cited third party test, not our own.
Setup
For a standard inbound use case, setup is a template, a prompt edit, a calendar connection and a phone number. The 15 minute quick start in the docs is realistic for a demo agent. A production agent takes longer: you will write and rewrite the prompt, connect tools, set transfer rules to a human, and test edge cases like silence, interruptions and voicemail. Plan a week of iteration for anything customer facing.
Who it is for
- Small and mid sized businesses that want a no code builder but may grow into the API.
- Healthcare and other regulated teams that need HIPAA without a large add on.
- Inbound support, booking and intake lines with up to 20 simultaneous calls.
- Agencies building agents for clients on GoHighLevel, Make or n8n.
Who should skip it
- Teams that want one bundled rate. Bland charges a single connected minute rate with models included.
- The most cost sensitive high volume lines. Telnyx states about $0.06 per minute all in with 500 concurrent calls.
- Developers who want any provider at cost. Vapi passes models through at cost and lists more providers per layer.
- Anyone who wants a done for you service. PolyAI or an AI receptionist product will do the building.
Reviews
Retell holds a 4.9 Trustpilot score from 810 reviews with 94% five stars. Reviewers praise how fast they could build and how natural the voices sound. The recurring complaints are latency pauses on some calls, slower support replies and limits on outbound calling to some countries.
Verdict
Retell AI earns the top spot because it is the most complete self serve platform with the clearest pricing. Model your bill on the full stack, not the $0.055 headline, and choose the smallest LLM that works. If you need help deciding between it and the closest rivals, read Retell AI vs Vapi and Bland AI vs Retell AI, or see the full best AI voice agents ranking.
How Retell AI scores
- Value
- 4.5
- Features
- 4.6
- Ease of setup
- 4.5
- Integrations
- 4.4
- Trust
- 4.7
Pros and cons
Pros
- Every component is priced on one public page, so you can model cost before you sign up
- 20 concurrent calls included free, where Vapi includes 4 and ElevenLabs Pro includes 20 only on a $99 plan
- SOC 2 Type II, HIPAA, GDPR and ISO 27001 listed with no compliance surcharge
- Drag and drop flow builder plus testing and QA tools, and a quick start that gets a phone agent live in about 15 minutes
Cons
- The $0.055 headline excludes the LLM, the voice and telephony, so quotes based on it understate the bill
- Premium voices and add ons stack: ElevenLabs v4 is $0.10 per minute and AI QA another $0.10
- Language coverage is not listed as a number; you check the dashboard
- Some Trustpilot reviewers report 1.5 to 2.5 second pauses and slow support replies
Retell AI plans and pricing
| Plan | Price | What you get |
|---|---|---|
| Pay as you go | $0 per month | $10 free credits, 20 concurrent calls, $0.055 per minute voice infra plus LLM, voice and telephony |
| Enterprise | Custom | Custom pricing and terms |
What users say
Retell AI reviews from users
No user reviews yet
Be the first to review Retell AIRetell AI alternatives
Intercom Fin4.6
Telnyx Voice AI Agents4.3
Vapi4.3
ElevenLabs Agents4.2
Ada4.0
Bland AI4.0Frequently asked questions
How much does Retell AI cost per minute?
The voice infra rate is $0.055 per minute, which includes speech to text. Add an LLM ($0.0032 to $0.16 per minute), a voice ($0.015 to $0.10) and telephony ($0.015 per minute on US Twilio). A budget stack totals about $0.0946 per minute and Retell's homepage quotes $0.07 to $0.31 all in.
Does Retell AI have a free plan?
There is no monthly fee and new accounts get $10 in free credits. You pay only for minutes and add ons after that, and 20 concurrent calls are included.
Is Retell AI HIPAA compliant?
Retell lists HIPAA alongside SOC 2 Type II, GDPR and ISO 27001, with no separate compliance fee on its pricing page. Sign a BAA with Retell before handling patient data.
What is Retell AI's latency?
Retell's homepage states about 600 ms. Some Trustpilot reviewers report pauses of 1.5 to 2.5 seconds, which usually comes down to the LLM, voice and prompt length you choose.
Can Retell AI make outbound calls?
Yes. Retell supports outbound and batch calling, with batch calls adding $0.005 per dial on top of the normal minute rate.
What integrations does Retell AI have?
Telephony through Twilio, Vonage or any SIP trunk, plus HubSpot, Cal.com, Make, n8n and GoHighLevel, and custom APIs as agent tools.
Is Retell AI better than Vapi?
For most businesses, yes, because of the visual builder, 20 free concurrent lines and included HIPAA. Vapi is cheaper per minute at the low end and gives developers more provider choice. See our Retell AI vs Vapi comparison for the math.