Bland AI pricing and review

I run Reachkit, a cold email tool rather than a voice platform; both are ways to reach prospects. I checked Bland’s prices on October 15, 2026, on its own pricing page and billing docs.

Bland AI pricing has two self-serve plans billed by the minute, plus Enterprise on a contract. These figures are from bland.ai/pricing as of October 15, 2026:

Talk time covers the language model, speech recognition and voice, billed to the second. The billing docs add a $0.015 minimum per outbound attempt on Bland’s lines, failed calls included, and $0.02 per text (the pricing page lists SMS under Enterprise). Extra numbers cost $15 a month. Transfer minutes apply only on Bland’s numbers.

Where Bland’s pages disagree. The billing docs list a Scale plan ($499 a month, $0.11 a minute) and a $29.99 Agent Phone plan for one agent and one US number, neither shown on the pricing page. The homepage says the per-minute rate covers telephony; the pricing FAQ says telephony is billed separately, at pass-through cost on Bland’s lines.

The rate went up. Until December 5, 2025, every self-serve plan paid $0.09 a minute, so pages still quoting $0.09 are out of date.

What a typical setup costs (my math from list prices, before any separate phone charges): an inbound line taking 500 minutes a month on Start runs about $70. Build at 5,000 minutes is $899. Calling 3,000 leads a month, with 40% answering for two minutes and $0.015 per unanswered dial, is about $363 on Start, right at its 100-calls-a-day cap, or about $614 on Build.

Bland builds AI phone agents that hold full conversations: answering inbound lines, calling people back, booking appointments, verifying details and taking payments. Bland says it runs its own speech recognition, language model and voice rather than renting them, in more than 40 languages.

Since September 2026, agents typically start from a prompt in a new Bland Agents builder. Older node-based Conversational Pathways must move over by November 15, 2026, per the migration docs. Agents use knowledge bases, webhooks and custom tools, can be tested against other agents before launch, and AI judges grade past calls.

Agents run on Bland numbers or your own Twilio or SIP line, and outbound calls go out singly or in batches through the API. A hosted MCP server lets AI assistants like Claude and ChatGPT use Bland, and contact memory syncs to HubSpot and Salesforce.

Strengths

• One rate for the AI. Talk time includes the model and voice, with “No token charges” (pricing), which is easier to forecast than separate providers.
• Security paperwork. The pricing FAQ lists SOC 2 Type I and Type II, PCI DSS and HIPAA eligibility with a signed BAA, and Bland announced FedRAMP 20x Class A in September 2026.
• Volume and backing. Fortune reported in June 2026 that Bland handles more than 3.5 million calls a week for 250-plus enterprise customers.
• Room for developers. An API, GitHub versioning, evals and agent tests, with frequent changelog releases.

Drawbacks

• Higher rates, many meters. Self-serve minutes rose from $0.09 to between $0.11 and $0.14 in December 2025, and attempts, transfers, numbers and texts are billed on top.
• Tight limits before $299. Start allows 100 calls a day and 10 at once.
• Much of it is Enterprise. Per the pricing page, SMS, warm transfers, guardrails, the scheduling node, SSO and Slack support need Enterprise, which has no public price.
• A learning curve. Summaries of G2 reviews say the platform can be complex and hard to learn.

Bland has few public reviews for its size. G2 shows 5.0 from 11 reviews per search results in October 2026, a small sample whose summaries praise quick deployment, low latency and ease of use once learned. Capterra lists no reviews, Trustpilot shows two on a profile Bland hasn’t claimed, and Product Hunt has two from about three years ago.

The top Reddit result for “bland ai review” in October 2026 is one tester’s comparison of voice agents from about a year earlier, which reports for Bland “almost no traction, people hung up”. That’s a single test from before the 2026 builder and speech releases, not a pattern.

Bland fits enterprises and contact centers in healthcare, insurance and finance with lots of calls where mistakes are costly: replacing a phone menu, covering after hours, scheduling, verification and collections. A signed BAA and on-prem hosting are available on Enterprise.

Look elsewhere if you hoped an AI would cold-call a bought list. Start stops at 100 calls a day, and Bland’s terms make you responsible for complying with all applicable laws and prohibit “unauthorized spam or bulk communications”. In outbound, I’d use email to find out who wants a conversation, and save the phone agent for the people who replied.

• Retell AI: voice agents with no platform fee, from $0.07 a minute, with model, voice and phone line priced as separate parts. See the Retell AI profile.
• Vapi: a developer platform at $0.05 a minute for hosting, where you pick and pay for each provider on top. See the Vapi profile.
• Close: a sales CRM with calling built in and Chloe, an AI agent that calls and qualifies leads, from $19 a user monthly. See the Close profile.
• Artisan: Ava, an AI SDR for email outreach that doesn’t place calls herself, from $280 a month. See the Artisan profile.
• Reachkit (our product): if you wanted Bland to reach cold prospects, Reachkit does that by email. Describe the campaign in a sentence and it finds verified leads (1,000 credits a month on the $39 Growth plan), writes the sequence, sends from warmed inboxes and labels replies like Meeting Requested. It doesn’t call or answer replies for you.