Last updated: 2026-11-04
Bland AI is an enterprise voice AI platform that builds phone agents on its own telephony, speech and orchestration stack. The vendor reports sub-400 ms latency, capacity for 1 million concurrent calls and support for 40+ languages. It lists SOC 2 Type II, ISO 27001, HIPAA and FedRAMP 20x Class A status for regulated buyers.
About Bland AI
Healthcare networks, insurers, banks and logistics operators use Bland to put software on the phone. The product lets a team build agents that place and answer calls, and the vendor says it owns every layer under them: the phone lines, transcription, the language model, the voice and the pathway logic that steers a conversation. A top-10 US bank is among its listed customers, and a careers page describes the company as Series C.
The core idea is billing and control in one place. Where Vapi and Retell AI charge a platform rate and then pass through the language model, speech and carrier costs, Bland quotes a single per-minute price that covers the model, transcription and voices, while phone-line charges are extra, either at Bland's carrier cost or via your own provider. Conversations are built as pathways: ordered steps with edge labels, loops and global steps, so a call follows a defined route instead of a free-form prompt. Agents can call tools and webhooks mid-call, transfer to a person, and send texts during the call.
It fits teams that need a phone agent in production inside a compliance review: appointment reminders, claims intake, collections, dispatch and inbound routing. Developers can start on the self-serve Start plan and send a first test call from the REST API the same afternoon, while enterprise buyers get a forward deployed engineer team, on-premises or VPC deployment, a signed BAA, SSO and data residency. Hokai's best AI receptionist guide and best AI voice assistants guide show where phone agents of this kind sit against lighter products such as NextPhone and echowin.
Pricing has three levels: a no-card Start plan, a Build plan with a monthly platform fee, and Enterprise contracted to volume. The entry plan limits parallel calls and daily volume tightly, and Build lifts both ceilings several times over. Warm and live transfers, guardrails, SMS and web chat, BAA and SSO sit on the Enterprise column of the plan table.
The vendor publishes a hosted MCP server and a CLI, so a coding agent can manage calls and pathways directly. Its speech layer pairs with the same family of products as ElevenLabs for voices and Deepgram for transcription, although Bland bundles its own so those are optional rather than required. For buyers comparing the build-it-yourself route against a managed one, the trade is control of each component against one invoice and one vendor to call when something fails. Teams whose phone work is mostly customer service may also look at Sierra or Decagon, and developers weighing another voice-agent builder can compare Bolna AI.
Pricing
Start: $0.14 per minute, no platform fee, no card, 2 credits and an inbound number on signup. Build: $0.12 per minute plus a $299 monthly platform fee. Enterprise: custom, contracted to volume.
Transfers bill at $0.05 per minute on Start and $0.04 on Build. LLM, speech-to-text and text-to-speech are included in the per-minute rate; phone-line charges come on top, at Bland carrier cost or through your own provider. Source: bland.ai/pricing, read 2026-11-04.
| Tier | Monthly price | What it includes |
|---|---|---|
| Start | Free | |
| Build | $299/mo | |
| Enterprise | Custom |
Key Features
- Conversational pathways: Calls follow ordered steps with edge labels, loop conditions and global steps, so agents stay on a defined route through regulated scripts.
- All-in-one per-minute rate: A single per-minute rate covers the language model, transcription and voices, with no token charges.
- Own telephony and SIP trunks: Numbers, inbound and outbound SIP trunks with failover and live traces run on Bland, or you can bring your own Twilio account.
- Voice cloning and live translation: Clone a voice from about ten seconds of audio and translate speech across 23 languages, including during transfers.
- Tools, webhooks and post-call data: Agents call external APIs mid-call, and signed post-call webhooks return transcripts and structured outcomes.
- Hosted MCP server and CLI: A hosted MCP server and a command-line tool let coding agents manage agents, calls and pathways without the dashboard.
- Enterprise deployment options: Top-tier contracts add self-hosted options, single sign-on, regional data storage and an embedded engineering team.
Pros
- One per-minute rate includes the language model, transcription and voices, so there are fewer vendor invoices to reconcile than on a Vapi or Retell stack.
- Pathways give a fixed call route with loops and global steps, which suits scripted work such as claims intake and appointment confirmation.
- The vendor lists SOC 2 Type I and II reports, ISO 27001, HIPAA with a BAA and GDPR with a DPA, plus FedRAMP 20x Class A, which shortens security review for healthcare and finance buyers.
- Self-serve signup on the Start plan needs no card and the vendor positions the entry plan as a quick start for developers.
Cons
- Warm transfers, live transfers, guardrails, SMS and web chat and the BAA sit on the Enterprise column, so small regulated teams cannot test them on Start or Build.
- Telephony is billed on top of the per-minute rate, so the true cost per minute depends on your carrier choice.
- The Start plan caps at 100 calls a day and 10 concurrent calls, which blocks load testing before an upgrade to the Build tier.
- The vendor markets to enterprise buyers, so hobby builders get less community material than with developer-first rivals.
Data Handling
- Compliance
- SOC 2 Type I and II · ISO 27001 · HIPAA with BAA · GDPR with DPA · FedRAMP 20x Class A · PCI DSS (self-attested)
Frequently Asked Questions
How much does Bland AI cost per minute?
The Start plan charges $0.14 per minute with a $0 platform fee, and the Build plan charges $0.12 per minute plus $299 a month. Transfer minutes bill at $0.05 on Start and $0.04 on Build. Enterprise is custom. Telephony is billed separately, at pass-through cost or on your own carrier.
Can you try Bland AI without paying?
Signing up on the Start plan gives you 2 credits and an inbound number, and the vendor says no card is required. After the credits run out you pay per minute, so it is a trial rather than a permanent free tier. The Start plan also caps you at 10 concurrent calls and 100 calls a day.
What are the main alternatives to Bland AI?
Vapi suits teams that want to pick their own model, voice and carrier. Retell AI is another developer platform that prices components separately. For a small business that wants a phone agent without a build project, look at NextPhone.
How does Bland AI compare with Vapi on cost?
Vapi and Retell advertise a platform rate and bill the language model, speech and telephony on top, so the final per-minute figure depends on the providers you pick. Bland states its rate already includes the model and speech, which makes a bill easier to predict. The vendor publishes its own comparison, so check it against your own call mix.
How do you get started with Bland AI?
Create an account on the Start plan, buy or port a number, and place a first call through the REST API or the dashboard. The vendor says most teams deploy a first agent within a day. Enterprise rollouts follow a 28-day framework with a forward deployed engineer.
Top Alternatives
- Vapi: Pick Bland if you want one bundled per-minute rate; choose Vapi when you want to swap model and voice providers yourself.
- Retell AI: Pick Bland when scripted pathways and enterprise deployment matter; choose Retell when you want a developer-first platform with pay-as-you-go components.
- NextPhone: Pick Bland for high-volume enterprise call flows; choose NextPhone for a lighter small-business phone agent.