Retell AI is a voice AI provider that runs each call on language and transcription models from other companies, with a low advertised starting rate. This review covers how Retell works, what a minute costs, what its contract says on uptime and medical calls, and why enterprises often choose Bland, as of November 2026.
What is Retell AI and how does it work?
Retell AI is a voice agent platform for AI phone agents, where a customer chooses most of the parts of a call and Retell runs them together. Retell sells voice infrastructure at $0.055 per minute, and everything else on a call is assembled on top of that.
A call on Retell moves through four layers. Deepgram, Azure, Soniox, or AssemblyAI turns a caller's speech into text, and a language model the customer picks, such as GPT, Claude, or Gemini, decides what the agent says. Retell sells its own platform voices next to voices from companies such as ElevenLabs, Cartesia, MiniMax, and OpenAI, and calls run over Retell's Twilio or Telnyx lines or over a customer's own carrier.
How easy is Retell AI to use?
A team builds call logic on Retell in one of two ways: a single prompt runs the whole call, or Conversation Flows split it into nodes. Retell sells flows for "precise call control" and warns that they "can make conversations harder to handle when callers give information early or change their minds", so the option with the most control also carries Retell's own warning. A team can start without a contract, though enterprise terms still sit behind a sales call, and each model or voice it picks is one more vendor to review and track. Retell also doesn't offer iMessage agents, so its agents reach customers by phone and SMS.
How does one call run on Bland and on Retell?
A call on Bland stays with one AI company, while a call on Retell can touch up to 4 AI companies, counting Retell.
1 AI company on each call
$0.14 a minute for all of its AI on self-serve Start, no platform fee
- TranscriptionDeepgram, Azure, Soniox, or AssemblyAI
- Language modelGPT, Claude, or Gemini
- SpeechElevenLabs, Cartesia, Minimax, or OpenAI
Up to 4 AI companies on each call, counting Retell
$0.14 a minute before phone, using GPT-4.1 and an ElevenLabs Flash voice
How much does Retell AI cost?
Retell AI advertises $0.07 to $0.31 per minute as of November 2026, and one production setup with GPT-4.1, an ElevenLabs Flash voice, and Twilio phone comes to $0.155 per minute before add-ons. Before phone, that setup is $0.14 per minute, and the bill runs across separate lines for voice infrastructure, a language model, and a voice.

The advertised range hides how far a bill can travel. One language model alone lists at $0.384 per minute, and once a prompt passes 4,000 tokens Retell scales billed time, so a 12,000-token agent bills a 60-second call as 180 seconds. Calls above a customer's concurrency limit cost an extra $0.10 per minute for the whole call.
Add-ons sit on top of that. PII Removal adds $0.01 per minute, Branded Call adds $0.10 per outbound call, and knowledge bases, Safety Guardrails, and Advanced Denoising add $0.005 per minute each.
How is Bland priced?
On self-serve plans, Bland Start is $0.14 per minute with no platform fee, and Bland Build is $0.12 per minute with a $299 monthly platform fee, which is lower than Start above about 14,950 minutes a month. Either rate covers the language model, transcription, and speech with no token charges, and phone is billed separately at carrier cost. Enterprise pricing is custom, set with Bland's sales team. Enterprise is the plan regulated teams buy. It adds native iMessage agents alongside SMS and web chat, warm transfers, guardrails, a BAA, SSO, data residency, dedicated orchestration with dedicated GPUs run by Bland, and forward-deployed engineers who work beside a customer's team, with concurrency contracted to its volume and custom pricing.
| Vendor | What a minute pays for | Total per minute | Phone |
|---|---|---|---|
| Bland Start (self-serve) | All AI: model, transcription, speech $0.14 | $0.14 before phone | Billed separately at carrier cost |
| Retell | Voice infrastructure $0.055, GPT-4.1 $0.045, ElevenLabs Flash voice $0.04 | $0.14 before phone | Twilio at $0.015 a minute, billed on top |
Bland Start (self-serve): Self-serve Start has no platform fee. Self-serve Build is $0.12 a minute plus $299 a month. Phone and transfers bill separately. Retell: Retell Voice Infra, GPT-4.1, and an ElevenLabs Flash voice, before Twilio phone minutes.
How does Retell AI perform?
Retell AI describes its latency as sub-second, and it quotes "around 600ms" on its own pages, while Bland publishes sub-200ms. Performance on Retell also depends on the outside companies a customer picks, because the lab behind a chosen model controls how it behaves and how long it stays available.
An outage works the same way. Retell says its fallback covers language models and voices, though a fallback voice comes from a different voice provider, so a caller can hear a different voice during an outage. Its own warning about Conversation Flows covers the moments callers change their minds, which is when a call is hardest to get right.
Support is another place where Retell's marketing and its details part ways. Retell lists "24/7 Support", then describes a "24h SLA, with active support between 9 AM and 9 PM PST (lower for enterprise customers)".
Bland sends alarms on latency, long silences, transcription quality, and API errors by email, phone call, or API, and its evaluations grade call audio as well as the transcript.
Which sounds more human, Retell AI or Bland?
Bland Speech v3 ranked 1st of 13 AI voices at 1365 on Design Arena's Audio Realism Bench on 4 August 2026, and Retell's own platform voices didn't appear in those results.
- #1BlandBland Speech v31365
- 2ndMicrosoftMAI-Voice-21224
- 3rdxAIGrok TTS1156
- 4thOpenAI, on RetellGPT Realtime 21119
- 5thGoogleGemini 2.5 Pro TTS1109
- Not rankedRetell's own voicesNot ranked
Ranks are among 13 AI voices, and Retell's own voices weren't ranked. Real human recordings scored 1500, at the right edge of this scale. Bars start at 600, not zero. Source: Design Arena's published results.
In Design Arena's blind test, vetted native speakers hear two voices read one script and choose the one that sounds more like a real person. The highest outside voice a Retell customer can choose, OpenAI's GPT Realtime 2, ranked 4th at 1119. See Design Arena's published results.
Is Retell AI secure enough for enterprise calls?
Retell AI reports HIPAA, SOC 2 Type II, ISO 27001, and GDPR compliance, but it stops short in places enterprise buyers ask about. Retell says its services are "not designed for use" in medical, emergency, or other safety-critical environments, even though it markets "HIPAA-Compliant Patient Calls". Customers can self-sign a BAA, the contract HIPAA requires before a vendor handles patient information, and until one is signed Retell's terms treat patient information as "Prohibited Data".
Card payments and uptime raise the same question. Retell bars card data unless it agrees in writing, and it advertises "We guarantee >99.9% uptime" while its contract promises "commercially reasonable efforts" at 99.5%, excluding up to six hours a month of scheduled maintenance.
Retell's language model, speech recognition, and text-to-speech providers are subprocessors, so a security review grows with each one.

How does Bland handle security?
Bland is full stack, so no outside AI lab sits in its call path. It includes a BAA on Enterprise, and its security program self-attests to PCI DSS v4.0.1 and holds FedRAMP 20x Class A. Enterprise customers get dedicated orchestration and dedicated GPUs run by Bland. Bland also offers a 99.9% uptime SLA on every plan.
Bland also runs its own SIP signaling and media servers, so no third-party telephony provider sits in the call path, and guardrails run through the whole call, with four TCPA guardrails built in: AI disclosure, recording disclosure, self-introduction, and opt-out.
Is Retell AI worth it for enterprises?
For most enterprise calls, Retell AI isn't the best choice as of November 2026, because several companies sit behind each call and its contract promises less than its marketing. Retell suits a small engineering team at modest call volume that wants to choose and pay for its own model vendors.
Retell AI pros and cons
Pros
- Customers can pick a language model and voice, and each pick is another vendor to review and track
- Retell charges no platform fee, and a bill runs across several separate lines
- A team can start without a contract, and enterprise terms still sit behind a sales call
Cons
- 99.5% uptime in the contract against more than 99.9% in its advertising
- "Not designed for use" in medical or safety-critical environments
- No PCI status, with card data barred unless Retell agrees in writing
- Long prompts scale billed time, and add-ons sit on top
- Language and transcription models come from other companies on every call
- No iMessage agents
Why Bland is the Retell alternative for enterprises
Bland is often the choice for enterprise customers, with more than 250 enterprise customers and over 1.3 billion calls handled. It builds and runs its language, speech, and transcription models, and its language model is a closed-source model it developed itself, so a security team reviews one AI company and one published rate covers all of that AI.
On Enterprise, forward-deployed engineers work beside a customer's team to launch its first agent. Enterprise customers get dedicated orchestration and dedicated GPUs run by Bland. Bland also runs branded iMessage agents natively on Enterprise, on the same pathways as voice. Bland's voice ranked first among AI voices on Audio Realism Bench on 4 August 2026, and Bland vs Retell has the full comparison.
How do Retell AI and Bland stack up at a glance?
This table sets out the rows enterprise buyers ask about first.
What these rows mean
- AI on each call: This row shows whether the language model, speech, and transcription behind each call belong to the platform or to other companies: Full stack means all three are the platform's own, Outsourced means its language model and transcription come from other companies, and Mixed means it runs some of its own models alongside outside ones. Bland builds and runs its own language, speech, and transcription models, so a call never passes through an outside AI provider.
- Audio Realism Bench score, 4 August 2026: Audio Realism Bench is Design Arena's blind listening test, where listeners pick whichever of two voices sounds more like a real person, and a higher score means listeners picked that voice more often. In its published results of 4 August 2026, Bland Speech v3 ranked first among AI voices. Each rival number is the highest-scoring voice model in those results available on that platform.
- iMessage: iMessage lets an agent text customers on a channel they already use. Bland offers it natively on Enterprise.
- HIPAA with a BAA: A Business Associate Agreement, or BAA, is the contract that lets a vendor handle patient information under HIPAA. Bland includes its BAA on Enterprise.
- PCI DSS: PCI DSS is a security standard from the card industry for handling card details. Bland self-attests to v4.0.1 and shares its attestation under NDA on Enterprise. Retell gives no PCI status, and it bars card data unless it agrees in writing.
- FedRAMP: FedRAMP is the US government's cloud security program. Bland holds 20x Class A.
- AI rate per minute: Headline rates cover different things, so a buyer should check what each one includes before comparing them. Bland's self-serve rate covers all of its AI: language model, transcription, and speech. Phone is billed separately at carrier cost, and Enterprise pricing is custom, set with Bland's sales team. Retell advertises $0.07 to $0.31 a minute, and one language model alone lists at $0.384 a minute.
A demo built on a real customer workflow is the quickest next step, and Retell AI alternatives and Bland as a Retell alternative go further.
Frequently asked questions
How much does Retell AI cost per minute?
Retell AI advertises $0.07 to $0.31 per minute, built from separate lines for voice infrastructure, a language model, a voice, and phone. With GPT-4.1, an ElevenLabs Flash voice, and Twilio phone it comes to $0.155 per minute before add-ons.
Is Retell AI HIPAA compliant?
Retell AI reports HIPAA compliance and customers can self-sign a BAA, though Retell says its services are "not designed for use" in medical, emergency, or other safety-critical environments. Bland includes a BAA on Enterprise, with no outside AI company to review.
Is Retell AI PCI compliant?
Retell AI bars card data unless it agrees in writing. Bland self-attests to PCI DSS v4.0.1.
What uptime does Retell AI guarantee?
Retell AI advertises "We guarantee >99.9% uptime", while its contract promises "commercially reasonable efforts" at 99.5%, excluding up to six hours a month of scheduled maintenance. Bland offers a 99.9% uptime SLA on every plan.
Does Retell support iMessage?
Retell doesn't offer iMessage agents as of November 2026. Bland runs branded iMessage agents natively on Enterprise, which send and receive photos, videos, files, voice memos, and tapbacks, and share the same pathways as its voice agents.
Does Retell AI charge for PII redaction?
Yes, Retell AI's PII Removal costs an extra $0.01 per minute. Bland redacts PII in real time on recordings and transcripts.
Is Retell AI good for enterprises?
Retell AI suits teams that want to manage their own model vendors. Enterprises that need one AI company on each call or the most realistic voices are better served by Bland.
