Bland AI API
★ Only Publicly Available OpenAPI DocumentCommunicationsVoice Telephonybearer42 EndpointsREST
For Agents
Place programmatic outbound phone calls with conversational AI agents, manage call pathways and knowledge bases, send SMS, and analyse call recordings and transcripts.
Use for: Place an AI phone call to a customer, Send an SMS through Bland AI, Get the recording for a completed call, Retrieve the corrected transcript of a call
Not supported: Does not handle video calling, in-app push notifications, or carrier-grade SMS shortcodes - use for AI phone calls, SMS, and conversational voice agent management only.
Jentic publishes the only available OpenAPI specification for Bland AI API, keeping it validated and agent-ready. The Bland AI API powers programmatic phone calls and conversational voice AI: place outbound calls, run them through configured agents and pathways, capture recordings and transcripts, and analyse call outcomes after the fact. The spec also covers batch calling, inbound number purchasing, voices and personas, knowledge bases, custom tools that an agent can call mid-conversation, contact management, SMS, and blocked-number controls. It is the full stack for building AI voice agents that can hold a real phone conversation.
Install Jentic One Beta
Connect the Bland AI API to your agent
Jentic One is a self-hosted execution layer for AI agents. It lets your agent call the Bland AI API, or any other public or private API you need. You set the rules, the agent never sees your credentials, and every call is logged.
Two steps, two machines. Install the instance in a safe environment, then register your agent from wherever it runs.
1
Step 1: Jentic One Host machine
# On the machine that will host your Jentic One instance:
curl -fsSL "https://jentic.com/install.sh?src=apis&api=%2Fapis%2Fbland.ai%2Fbland" | sh2
Step 2: Agent machine
# On the machine where your agent runs (keep this separate from the instance):
curl -fsSL "https://jentic.com/install.sh?src=apis&api=%2Fapis%2Fbland.ai%2Fbland" | sh
jentic register # connects your agent to your Jentic One instanceJentic One is in public beta. The setup above keeps your agent separate from the instance, which is what you want before using real credentials: an agent running as the same OS user as Jentic One can read its stored keys directly. Just evaluating? A single local install is fine to start. See the secure deployment guide for the tiers.
Capabilities
What an agent can do with Bland AI API.
Place an outbound AI phone call with a configured agent and pathway
Stop, list, and inspect calls in flight, including recordings and corrected transcripts
Run batch calling campaigns with /batches and stop them mid-flight
Configure conversational agents, pathways, voices, personas, and knowledge bases
Define custom tools the agent can invoke during a live call
Send transactional or batch SMS and manage blocked-number lists
Use Cases
Patterns agents use Bland AI API for, with concrete tasks.
★ AI Outbound Calling Campaign
Run an outbound calling campaign that places personalised AI phone calls to a contact list and routes responses through a configured pathway. POST /batches creates the campaign, GET /batches/{batchId} reports progress, and GET /calls/{callId}/recording plus POST /calls/{callId}/analyze surface the outcome of each call. The same flow is reused for appointment reminders, lead qualification, and customer outreach.
POST /batches with the contact list and target pathway, poll GET /batches/{batchId} until complete, then call POST /calls/{callId}/analyze on each call and report the outcomes.
Live Conversational Voice Agent
Build a conversational voice agent that can hold an open-ended phone call: configure the agent via POST /agents, attach knowledge through /knowledge/learn/text or /knowledge/learn/web, and define tools the agent can invoke mid-call via POST /tools. When a call is placed with POST /calls, the agent uses the pathway and tools to answer questions, take actions, and follow up with SMS via POST /sms/send.
POST /agents to create the agent, POST /knowledge/learn/text to attach the FAQ, POST /tools to expose a CRM lookup, then POST /calls to place a call routed through the agent.
Post-Call Analysis and Transcript Review
After a call completes, pull the recording and corrected transcript to feed into QA, compliance review, or downstream analytics. GET /calls/{callId}/recording returns the audio, GET /calls/corrected-transcript returns the cleaned transcript, and POST /calls/{callId}/analyze produces structured outcome metadata. This loop turns Bland AI calls into a measurable funnel.
Call POST /calls/{callId}/analyze for each completed call, then GET /calls/corrected-transcript and store both in the QA database.
Inbound Number Setup and Routing
Provision a Bland AI inbound number to receive calls, attach a pathway and agent, and route the inbound conversation accordingly. POST /inbound/purchase buys the number, GET /inbound lists provisioned numbers, and the pathway+agent endpoints define how the inbound call is handled. This is the foundation for AI-handled inbound support, sales, or appointment lines.
Call POST /inbound/purchase to acquire a number, then POST /agents with the inbound pathway id so calls to the number are routed through the configured agent.
Agent-Driven Call Through Jentic
An AI orchestrator agent decides the user needs a call rather than an email. It searches Jentic for 'place an AI phone call', loads the POST /calls schema, and executes the call with the supplied phone number and prompt. Bland AI bearer credentials remain in your Jentic One instance, so the orchestrator never sees the raw API key. This makes Bland AI a drop-in voice tool for higher-level agent workflows.
Use Jentic search for 'place an AI phone call to a customer', load the POST /calls schema, then execute it with the phone number, agent id, and call prompt.
Key Endpoints
42 endpoints — jentic publishes the only available openapi specification for bland ai api, keeping it validated and agent-ready.
METHOD
PATH
DESCRIPTION
POST
/calls
Place an AI phone call
GET
/calls/{callId}
Get call details
POST
/calls/{callId}/analyze
Analyze a completed call
POST
/batches
Create a batch calling campaign
POST
/agents
Create a conversational agent
POST
/knowledge/learn/text
Add text to a knowledge base
POST
/sms/send
Send a single SMS
POST
/inbound/purchase
Purchase an inbound number
POST
/calls
Place an AI phone call
GET
/calls/{callId}
Get call details
POST
/calls/{callId}/analyze
Analyze a completed call
POST
/batches
Create a batch calling campaign
POST
/agents
Create a conversational agent
POST
/knowledge/learn/text
Add text to a knowledge base
POST
/sms/send
Send a single SMS
POST
/inbound/purchase
Purchase an inbound number
Why Jentic?
What agents get from Jentic-routed access to this vendor.
Setup
Wiring the Bland AI API by hand means building the bearer Authorization header, targeting the api.bland.ai/v1 host, and handling retries across call, SMS, and number-purchase endpoints yourself. Through Jentic you install once, import Bland AI from the API Directory, store the key once, and your agent calls it.
Permission scoping
Bland AI puts the call id in the URL path (/calls/{callId}), so a rule can pin your agent to reading and analyzing a specific call. You choose the operations it may call, so charge-incurring actions like POST /calls, POST /sms/send, and POST /inbound/purchase are not included unless you add them.
Credential isolation
Your Bland AI bearer key is stored once, encrypted, by your own Jentic One instance and injected into the Authorization header at execution time. It never enters the agent's prompt, logs, or context.
Intent-based discovery
Agents search Jentic by intent such as 'place an AI phone call to a customer' or 'send an SMS through Bland AI', and Jentic returns the matching operation with its input schema so the agent calls the right endpoint without browsing the reference docs.
Related APIs
Alternatives and complements available in the Jentic catalogue.
FAQs
Specific to using Bland AI API through Jentic.
Why is there no official OpenAPI spec for Bland AI API?
Bland AI does not publish an OpenAPI specification. Jentic generates and maintains this spec so AI agents and developers can call Bland AI via structured tooling. It is validated against the live API and kept up to date. Get started with Jentic One, the self-hosted execution layer.
What authentication does the Bland AI API use?
HTTP bearer authentication (security scheme bearerAuth). Callers attach a bearer token in the Authorization header. Through Jentic the token is stored in the encrypted vault, and the agent only ever sees a scoped Jentic token, not the raw Bland AI API key.
Can I place an AI phone call with the Bland AI API?
Yes. POST /calls is the entry point - supply the phone number, the agent id (or inline configuration), and the call task. Use GET /calls/{callId} to follow status and POST /calls/{callId}/stop to halt a call early.
How do I get the recording and transcript for a completed call?
Call GET /calls/{callId}/recording for the audio file and GET /calls/corrected-transcript for the cleaned transcript. POST /calls/{callId}/analyze returns structured outcome metadata you can store alongside the recording.
What are the rate limits for the Bland AI API?
The OpenAPI spec does not declare an explicit rate limit. Bland AI publishes per-account concurrency limits that scale with plan; design batch calling and SMS workflows to back off on HTTP 429 responses and respect the published per-minute quotas on your tier.
How do I place a Bland AI call through Jentic?
Install with pip install jentic, then use Jentic search with 'place an AI phone call to a customer'. Jentic loads the POST /calls schema; execute it with the phone number, agent id, and prompt. The bearer token never enters the agent's prompt context.
Is the Bland AI API free?
No. Bland AI is paid per minute of call time and per SMS, with rates published on the Bland AI website. There is no free tier in the API; account funding is required before /calls and /sms/send will succeed.
Can I limit what my agent is allowed to do with the Bland AI API?
Yes. Because you run Jentic One yourself, your own rules decide which Bland AI operations and credentials the agent may use. Since Bland AI puts the call id in the URL path (/calls/{callId}), you can pin an agent to read-only work like fetching recordings or running POST /calls/{callId}/analyze on a specific call, while charge-incurring operations such as POST /calls, POST /sms/send, and POST /inbound/purchase stay out of reach unless you explicitly add them. Your bearer key is injected only at execution time, so the agent can call only the operations you allow.
GET STARTED