Retell AI alternatives
Six voice agent platforms against Retell, with the real cost of a minute.
Short answer: Retell's $0.07 to $0.31 a minute is a sum of parts (voice infrastructure, model, voice, telephony, add-ons), and most teams leave when that sum, the support or the concurrency bill stops fitting. If you want natural voices and a monthly bundle you can predict, go with ElevenLabs Agents: Pro is $99 a month for 1,238 minutes. If your developers want every model at cost, take Vapi. If compliance and call quality come first, Telnyx runs the agent on its own carrier network with HIPAA in every plan.
Engineering teams building their own product should look at LiveKit, small businesses with no developer at Goodcall, and outbound teams at Bland AI. Below: why teams leave Retell, what each alternative costs for the same 1,000 minutes, and the add-on fees that turn a $0.07 minute into something else.
The ranking is editorial: a sponsored placement is always flagged on its card, and no affiliate link changes a score.
Why teams are leaving Retell AI
Retell AI is a platform for building AI phone agents: you write the agent's instructions, choose a language model and a voice, attach a knowledge base, and connect a phone number, and Retell runs the live conversation, transfers to humans, books appointments and analyzes each call afterwards. It is popular with developers and agencies because it goes live quickly on pay-as-you-go pricing with $10 in free credits. The reasons people leave are rarely about call quality. They are about how the per-minute bill is built, what support you get without an enterprise contract, and how much assembly the platform still asks of you.
The per-minute price is a sum, not a rate
Add-ons are billed per minute or per call
Concurrency costs $8 a line above 20
Support stays community and email until Enterprise
Custom contract terms wait for Enterprise
Phone numbers and countries add their own costs
ElevenLabs Agents
ElevenLabs started as a text-to-speech company and now sells a full agent platform, ElevenAgents: you describe the agent, connect a knowledge base, pick a voice, and plug it into a phone number. Retell itself resells ElevenLabs voices inside its own platform, so a team that chose Retell for those voices can go to the source.
How its price reads against Retell: Retell bills each minute by adding up its pieces, ElevenLabs sells bundles of call minutes. Pro costs $99 a month and includes 1,238 minutes with 20 concurrent calls; Creator is $22 for 275 minutes; Scale is $299 for 3,738 minutes. Paid yearly, Pro works out to $82.50 a month. Extra minutes cost $0.08 each, and if you exceed your concurrency limit, burst minutes cost $0.16, double the standard rate. Two lines sit outside the bundle: LLM usage is billed on top, based on the model you choose, and ElevenLabs adds no telephony fee because you connect your own provider, such as Twilio or a SIP trunk, which bills you for the carrier side.
Its limit: the bundle only beats Retell if you use it. The monthly fee is due in quiet months too, so a team with a very uneven call volume can end up paying for capacity it doesn't touch, where Retell would have billed nothing.
For the full picture of the platform, read our ElevenLabs review and the detail of every plan in ElevenLabs pricing. If it is not the fit either, our ElevenLabs alternatives page lists what teams move to.
- Call minutes bundled into monthly plans
- No limit on the number of agents
- 20 concurrent calls on Pro
- ✓Predictable monthly bill up to the bundle
- ✓Free plan to build and test
- ✓Bring your own telephony with no ElevenLabs fee on it
- ✗LLM usage billed on top of the bundle
- ✗Burst minutes at $0.16 when you exceed concurrency
- ✗Monthly fee due even in quiet months
| Criterion | ElevenLabs Agents | Retell AI |
|---|---|---|
| Pricing model | Monthly bundles of minutes | Pay per minute, by component |
| Entry | Free, 15 minutes | $10 in free credits |
| Concurrency | 4 free, 20 on Pro | 20 included, $8 each above |
| Telephony | Your own provider | Retell numbers or your own |
Pick ElevenLabs if voice quality sells your calls and you can size a monthly bundle. Take Vapi instead if your engineers want to swap every model and pay each one at cost.
Vapi
Vapi is the closest substitute to Retell for a developer: an API and a dashboard to orchestrate a voice agent, where you pick the transcriber, the language model and the voice, and Vapi runs the real-time call. The main difference is how the bill is built. Vapi charges $0.05 a minute for its hosting and passes the models through at cost, so you can see exactly what each provider charges.
How its price reads against Retell: on Vapi's own estimate for 1,000 minutes, hosting comes to $50, Deepgram transcription to $10, an OpenAI model to $8 to $45 and an ElevenLabs voice to $15 to $24, so $82 to $129 a month before telephony. Twilio inbound calls add $0.008 a minute, Telnyx numbers $0.0055. On Retell, a setup with platform voices and GPT 5.1 comes to $0.11 a minute, or $110 for the same 1,000 minutes. The two land in the same range; the difference is which levers you control.
Its limit: concurrency, support and compliance are sold separately. Without a package you get 4 concurrent calls, peer support on Discord and email support with no SLA. Core, at $29 a month, brings 10 concurrent calls and a two-business-day email SLA; Pro has a $999 monthly minimum; HIPAA compliance costs $2,000 a month; and each extra concurrent line is $10 a month. If you came from Retell's free 20 concurrent calls, count those lines before switching: 16 extra lines on pay as you go cost $160 a month. No-code teams can still trigger Vapi calls from Make, as our Vapi and Make integration guide shows. See what Vapi users move to in our guide to switching from Vapi.
- Models passed through at cost
- Choice of transcriber, LLM and voice per agent
- Telnyx, Twilio and Vonage telephony rates listed
- ✓Transparent per-component bill
- ✓$5 in credits to start
- ✓Deep API for custom flows
- ✗HIPAA compliance costs $2,000 a month
- ✗4 concurrent calls without a paid package
- ✗No support SLA below the $29 Core package
| Criterion | Vapi | Retell AI |
|---|---|---|
| Platform fee | $0.05/min hosting | $0.055/min voice infrastructure |
| Models | Passed through at cost | Priced per model per minute |
| Concurrency | 4 included, $10 per extra line a month | 20 free, $8 each above |
| HIPAA | $2,000 a month add-on | Standard BAA self-signed, custom on Enterprise |
Pick Vapi if your engineers want every component visible and swappable. Take ElevenLabs instead if you'd rather buy minutes in a bundle and stop tuning.
Telnyx
Telnyx is a telecom carrier, and that is its argument against Retell. Retell builds agents and connects them to Twilio or Telnyx numbers; Telnyx owns the network and runs the agent on top of it. Fewer hops between the caller and the model is the main lever on latency, and one vendor for both the line and the agent means one contract and one bill.
How its price reads against Retell: the voice engine, which bundles orchestration, speech-to-text and text-to-speech, costs $0.05 a minute, and Telnyx says a production agent lands near $0.06 a minute once LLM tokens and telephony are added (US inbound SIP from $0.0032 a minute, US local numbers $1 a month). Against Retell's $0.07 to $0.31 range, Telnyx starts below Retell's lowest price. HIPAA is included in every plan, pay as you go included, where Vapi sells it at $2,000 a month. The pay-as-you-go account handles up to 500 concurrent calls. Our Telnyx and Make integration guide shows how to connect it to the rest of your stack.
Its limit: Telnyx is built for engineering teams, and its cheap minute has two catches. Each call's voice engine time is rounded up to the next full minute, so a 15-second call bills as 60 seconds, where Retell tracks each call to the second. And pay as you go comes with community support: 24/7 support by chat or phone starts on Growth, a $2,000 monthly minimum spend. You also need a card on file to start.
- Carrier network and voice agent from one vendor
- HIPAA included in every plan
- 500 concurrent calls
- ✓Low per-minute price for production agents
- ✓HIPAA without a separate add-on
- ✓One bill for line and agent
- ✗Each call rounded up to the next full minute
- ✗Community support only on pay as you go
- ✗Card required to start
| Criterion | Telnyx | Retell AI |
|---|---|---|
| Per-minute price | $0.05 engine, about $0.06 in production | $0.07 to $0.31 a minute, pay as you go |
| HIPAA | Every plan | Standard BAA self-signed, custom on Enterprise |
| Concurrency | 500 concurrent calls | 20 included |
| Telephony | Own carrier network | Twilio or Telnyx numbers |
Pick Telnyx if compliance and call quality matter more than a visual builder. Take Goodcall instead if nobody on your team writes code.
LiveKit
LiveKit is real-time audio and video infrastructure with an agents framework on top. Where Retell gives you a finished agent platform, LiveKit gives you the building blocks: the media server, the agent sessions, telephony, and an inference layer that lets you pick the speech-to-text, LLM and voice models per minute. Its server and its Agents framework are open source under the Apache 2.0 license, so you can also self-host them. It is what you choose when the voice agent is your product rather than a tool your support team uses.
How its price reads against Retell: the Build plan is $0 a month with 1,000 agent minutes and no credit card, but also 5 concurrent sessions and $2.50 of inference credits, about 50 minutes of models: past that, you pay model prices or plug in your own provider keys. Ship starts at $50 a month with 5,000 agent minutes and Scale at $500 a month with 50,000, where HIPAA comes in. Agent sessions cost $0.01 a minute, inbound calls to a US local number $0.01 a minute (the number itself is $1 a month after the first free one); LiveKit's calculator shows a full stack (session, telephony, an open LLM, transcription, a voice and observability) at about $0.0479 a minute. That is well under Retell's $0.07 floor, if you are ready to build what Retell builds for you.
Its limit: past a simple agent, you write code. LiveKit's Agent Builder lets you prototype simple voice agents in the browser without code and download them as a Python project, but LiveKit says it is not meant to replace the Agents SDK, and booking flows or post-call analysis are built in code by your developers.
- 1,000 agent minutes free on Build
- Per-minute model prices listed in the calculator
- Agent sessions at $0.01 a minute
- ✓Per-minute stack well under Retell's floor
- ✓No credit card for the free plan
- ✓Open source, can be self-hosted
- ✗Real projects need developers
- ✗HIPAA starts on the $500 Scale plan
- ✗5 concurrent sessions on the free plan
| Criterion | LiveKit | Retell AI |
|---|---|---|
| Entry | Free, 1,000 agent minutes | $10 in free credits |
| Example per-minute cost | About $0.0479 | $0.07 to $0.31 a minute, pay as you go |
| HIPAA | From the Scale plan, $500/mo | Standard BAA self-signed, custom on Enterprise |
| Who builds | Your developers | Retell's builder or API |
Pick LiveKit if your team ships its own voice product. Take Vapi instead if you want the orchestration done and only the models to choose.
Goodcall
Goodcall is not a developer platform. It is an AI receptionist for small businesses: it answers the phone, qualifies callers, books work into your calendar and sends you the lead by text and email when the call ends. A plumber who used Retell's templates and gave up on the configuration is the person Goodcall is built for.
How its price reads against Retell: Goodcall doesn't bill by the minute but by unique customers, the number of different people who call you in a month. Starter is $79 a month for 100 unique customers, then 79 cents each, which Goodcall says is about 150 calls; Growth is $129 for 500, then 26 cents each; Scale is $299 for 2,000, then 15 cents each. Goodcall's own math puts Starter at about 13 cents a minute on four-minute calls. That is above Retell's floor, but the minute price stops mattering when nobody has to build the agent. Every plan includes a local number, and you can keep your existing one by forwarding it. If you are weighing it against a human service, see our ranking of telephone answering services.
Its limit: it is a receptionist, not a platform. Starter caps you at 5 workers, 1 external tool, 1 team member and 30 days of call history; Growth lifts that to 10 workers, 5 tools, 5 members and 90 days. Goodcall is built to answer your calls; for outbound campaigns, Bland or Retell fit better. Our page on AI receptionists compares the tools built for this job.
- Billed by unique customers, not minutes
- 90-day call history on Growth
- Ready-made receptionist, no flow to design
- ✓No code: Gabi builds the agent from your description
- ✓Repeat callers don't raise the bill
- ✓Clear overage price per customer
- ✗1 external tool and 1 team member on Starter
- ✗30-day call history on Starter
- ✗Overage of 79 cents a customer on Starter
| Criterion | Goodcall | Retell AI |
|---|---|---|
| Pricing unit | Unique customers a month | Minutes |
| Entry | $79 a month, 100 customers | $10 in free credits, then per minute |
| Setup | Ready-made receptionist | Builder or API |
| Audience | Small local businesses | Developers and contact centers |
Pick Goodcall if you want the phone answered, not a platform. Take ElevenLabs instead if you need several agents or languages.
Bland AI
Bland AI builds phone agents for enterprises and is known for outbound volume: reminders, qualification calls, follow-ups. Where Retell's headline is a range that depends on your LLM and voice, Bland quotes a single talk-time rate per plan, and its pricing page argues that the per-minute rates of Vapi and Retell are only the platform fee before the LLM, transcription and voice are added.
How its price reads against Retell: Start costs $0.14 a talk minute with a $0 platform fee and no card, and transferred minutes cost $0.05. Build drops talk time to $0.12 a minute but adds a $299 monthly platform fee. A thousand minutes on Start comes to $140, against $110 on Retell with platform voices and GPT 5.1: Bland is not the cheapest, it is the simplest to forecast. Telephony is billed separately, on your own carrier or Bland's at pass-through cost, and Bland lists SOC 2 Type I and Type II.
Its limit: Start allows 100 calls a day and 10 concurrent calls, which is tight for a real outbound campaign, so volume users end up on Build and its $299 fee, for 2,000 calls a day and 50 concurrent calls. Warm and live transfers are Enterprise features on Bland: on Start and Build, check that a plain transfer matches how your reps take calls. To connect Bland to a CRM without code, see our Bland and Make integration guide.
- One talk-time rate per plan
- SOC 2 Type I and Type II
- Transfer minutes priced lower than talk time
- ✓Simple to forecast per minute
- ✓No card to start
- ✓Built for outbound campaigns
- ✗100 calls a day on Start
- ✗Warm and live transfers on Enterprise only
- ✗$299 monthly fee to reach the lower rate
| Criterion | Bland AI | Retell AI |
|---|---|---|
| Per-minute price | $0.14 Start, $0.12 Build | $0.07 to $0.31 a minute, pay as you go |
| Platform fee | $0 on Start, $299/mo on Build | None on pay as you go |
| Daily limit | 100 calls a day on Start | 20 concurrent calls included |
| Compliance | SOC 2 Type I and II | SOC 2, HIPAA with a self-signed BAA |
Pick Bland if you run outbound volume and want one number per minute. Take Telnyx instead if you want a production price near $0.06 a minute.
6 Retell AI alternatives compared
Entry prices as each pricing page shows them. The edge column gives the one reason to pick each tool over Retell.
| Best for | Edge over Retell | Free plan | Team size | Visit | ||||
|---|---|---|---|---|---|---|---|---|
| 1 | ElevenLabs Agents | For voice quality on a fixed budget | Minutes bundled in monthly plans | 4.2/5 | Free, Pro $99/mo | ✓ | Small to mid-size teams | Visit → |
| 2 | Vapi | For developers | Every model billed at cost | 4.1/5 | $0.05/min + models | ✓ | Developer teams | Visit → |
| 3 | Telnyx | Carrier-grade option | Own carrier network, HIPAA in every plan | 3.8/5 | $0.05/min engine | — | Regulated industries | Visit → |
| 4 | LiveKit | Open infrastructure for builders | Free 1,000 minutes, full control | 3.8/5 | Free, Ship $50/mo | ✓ | Product engineering | Visit → |
| 5 | Goodcall | For small businesses without developers | Billed per unique caller, no build | 3.6/5 | Starter $79/mo | — | Local businesses | Visit → |
| 6 | Bland AI | For outbound call volume | One talk-time rate per minute | 3.5/5 | $0.14/min | ✓ | Outbound sales teams | Visit → |
Is your tool missing from this page? Email us: hello@hackceleration.com
You're leaving Retell because… go with
Start from what broke with Retell. Each card jumps to the tool that solves that problem best.
Minutes bundled into a monthly plan.
You want to see what each model costsVapiHosting at $0.05 a minute, models at cost.
You want the phone line and the agent from one vendorTelnyxOwn carrier network, HIPAA in every plan.
The voice agent is your own productLiveKitInfrastructure and agents framework, 1,000 minutes free.
Nobody on the team writes codeGoodcallAn AI receptionist built from a plain-English description.
You run outbound campaigns at volumeBland AIOne talk-time rate per minute.
Stay on Retell, or switch?
Staying makes sense if…
- Your volume is low or uneven, and paying per minute with no monthly fee costs less than any bundle.
- Your agents fit in 20 concurrent calls, so you never pay for extra capacity.
- You use Retell's call transfer, appointment booking and post-call analysis, which you would rebuild elsewhere.
- You have already tuned prompts and flows that work, and a migration would mean retesting every call path.
- You need a HIPAA BAA: Retell's standard one is self-signed at no extra fee.
Switch if…
- Your invoice per minute is closer to $0.30 than $0.07 once the model, voice and add-ons are counted.
- You need more than 20 simultaneous calls and the $8 lines add up.
- Your business can't run on community and email support.
- You need your own contract terms (a custom BAA, MSA or DPA), SSO or role-based access without an Enterprise negotiation.
How to choose a Retell AI alternative
A voice agent platform chains three models in real time. Speech-to-text turns the caller's words into text, a language model decides what to say and which action to take, and text-to-speech speaks the answer, all in well under a second so the conversation feels natural. Around that chain sit the telephony (the phone number and the carrier that carries the call), the tools the agent can use (a calendar, a CRM, a transfer to a human) and the analysis after the call. Some tools, like Retell, Vapi and Bland, sell the orchestration and let you choose the pieces. Others, like ElevenLabs, sell bundles. LiveKit sells the infrastructure for you to build on. Goodcall sells a finished receptionist. Platforms built for large contact centers, such as Sierra, PolyAI, Cognigy and Synthflow, which some lists include, sell their full agent platform through a sales conversation: Synthflow has no free plan, no trial and no rate by the minute, Sierra prices on outcomes, and PolyAI's self-serve offer is a model API at 2 cents a connected minute, without telephony or orchestration. If you need a full contact center suite rather than a voice agent, start from our ranking of call center software.
Before comparing prices, write down what you need: the countries you call, the number of simultaneous calls at your busiest hour, the compliance documents you must sign, and who will maintain the agent. Then demand four things from any vendor: a full per-minute cost for your actual model and voice, the concurrency included, the telephony price in your countries, and the support level on the plan you will buy. Then check how the agent will reach your CRM and calendar: our Make guides cover Retell, Vapi, Telnyx and Bland. For a broader view, see our ranking of the best AI voice agents and its versions for restaurants and real estate.
Minutes: the unit most tools bill
A minute is a minute of connected call time, but vendors count it differently. Retell tracks each call to the second, while Telnyx rounds each call's voice engine time up to the next full minute, so 1,000 calls of 15 seconds bill 250 minutes on one and 1,000 on the other. Retell, Vapi, Telnyx, LiveKit and Bland bill per minute, but each minute is a different bundle: on Vapi it is hosting only, with models added at cost; on Bland it is talk time with telephony on top. Compare the full price of a minute on your model, voice and call length, never the headline.
Bundles: minutes included in a plan
ElevenLabs sells call minutes inside monthly plans, with $0.08 per extra minute and $0.16 when you exceed your concurrency limit. Bundles beat per-minute billing when your volume is steady. Check the burst price, the LLM line billed on top, and whether telephony is yours to pay.
Unique customers: the receptionist unit
Goodcall counts unique customers, the different phone numbers that call you each month, not minutes. A regular customer who calls five times counts once, and spam and wrong numbers don't count at all. For a local business with returning callers, that unit is easier to budget than minutes, and the overage price per customer is listed on each plan.
Latency: the pause your callers hear
Latency is the silence between the caller finishing a sentence and the agent answering. In a chained setup, transcription, the language model and the voice each add to it, and the carrier adds the network on both ends. Speech-to-speech models skip steps but cost more: on Retell, GPT Realtime 1.5 is $0.345 a minute against $0.04 for GPT 5.1 in the chained setup. PolyAI quotes a sub-300 ms median time to first token for its Dialog-RSN-1 model. Ask each vendor for the median and the slow-call figure on your own model, voice and carrier, measured from your callers' region, then call the demo agent yourself from a mobile phone.
Concurrency and telephony: the fixed costs
Concurrent calls are the lines that can be busy at the same time. Retell includes 20 then charges $8 each a month, Vapi includes 4 without a package then charges $10 per extra line a month, ElevenLabs gives 20 on Pro. Telephony is the other fixed cost: a number a month plus a per-minute carrier rate, such as $0.0055 a minute for Telnyx numbers or $0.008 for Twilio inbound calls in Vapi. Add both to your per-minute math before you decide.
Retell AI alternatives: frequently asked questions
How long does it take to deploy a voice AI agent?
A first agent can be live within a day on a self-serve platform; an enterprise rollout with integrations, number porting and compliance review takes weeks. Bland, for instance, says most teams deploy a first agent within a day on its self-serve plans and runs enterprise deployments on a 28-day framework. Testing is the slow part: every call path needs to be called, listened to and fixed, which Retell's simulation testing and Goodcall's practice calls (about 100 test runs on Starter) speed up.
Are voice AI agents GDPR compliant?
They can be, but it depends on where audio and transcripts are stored and on the contract you sign. Ask each vendor for its data processing agreement, where its servers run, and whether you can turn off recording. Retell offers a self-signed DPA with EU Standard Contractual Clauses at no extra fee, per-agent retention from 1 day to 2 years and opt-out recording, but it does not currently operate services within the European Union, which matters if your data must stay there.
Can an AI voice agent book meetings directly into a calendar?
Yes. Retell lists appointment booking that checks availability and books, Goodcall books into a connected Google Calendar, and most platforms here connect to calendars through built-in tools or API calls. Check the plan, though: on Bland, the appointment scheduling node is an Enterprise feature. Test double bookings and time zones before going live.
Do voice AI platforms charge for hold time and transfers?
For hold time, usually yes; for transfers, it varies. Retell bills the whole call, silence and hold included, because its speech-to-text keeps listening, and once a call is transferred its agent fee stops and only telephony continues. Bland bills transferred minutes at $0.05 on Start and $0.04 on Build, and charges no transfer fees if you bring your own carrier. Get both rules in writing before you model a support line with long holds.
Can I keep my phone number when I switch from Retell?
If you bought the number through Retell's telephony, check with its support whether it can be ported before closing the account; if you connected your own number through a SIP trunk or a custom telephony provider, you only point that trunk at the new platform. Retell charges nothing for SIP trunking or custom telephony, which makes bringing your own number the safer setup from day one. Goodcall, for instance, lets you keep your existing number by forwarding it to the agent.
Do AI phone agents need consent to record calls?
Often, yes: recording-consent rules depend on where each party is, so the safest default is an agent that announces the recording in its greeting. Retell lets you opt out of recording and transcription, and choose per agent what it stores: everything, everything except personal data, or basic call attributes only.
What happens when more calls come in than my concurrency limit?
Depending on the platform, the extra calls are rejected, billed at a burst rate, or need more lines bought in advance. Retell rejects calls above your limit unless you turn on Concurrency Burst, which connects them for $0.10 a minute more; ElevenLabs connects them at its $0.16 burst rate; LiveKit's free Build plan stops at 5 concurrent sessions. Size the limit on your busiest hour, not on your monthly volume.
Can I run the same agent on several voice platforms to compare them?
Yes, and it is the best way to choose. Write the prompt and the tools once, rebuild the agent on two platforms, and route a share of calls to each through your own SIP trunk. Free credits go less far than they look: Retell's $10 buy about 90 minutes at $0.11, Vapi's $5 about 40 to 60 minutes at its $0.08 to $0.13 estimate, and LiveKit's 1,000 free agent minutes come with $2.50 of model credits. Plan a few dozen test calls on each, not hundreds.
Which Retell alternatives sign a HIPAA BAA, and at what price?
Most do, at very different prices. Telnyx includes HIPAA on every plan, pay as you go included; LiveKit signs a BAA from its Scale plan at $500 a month; Vapi sells HIPAA compliance, with a standard BAA and DPA, at $2,000 a month; Bland signs a BAA on Enterprise only. Retell itself lets you self-sign its standard BAA at no extra fee, so compliance alone is rarely a reason to leave it.