Bland AI connector for AI agents

Tools your voice ops agent reaches for on Bland AI, scoped per user.

CALL ANY TOOL

50 tools for voice agents in Bland AI: send and stop calls, run call batches, read transcripts and call analysis, edit pathways and agents, send SMS, and manage contacts, inbound numbers, and knowledge bases.

bland_analyze_call

Call analyze

Run AI analysis over a finished call transcript to answer questions you define. Returns status, message, an answers array in the same order as the questions, and credits_used. Use to extract structured answers from a call. Use get_call to read the raw transcript instead. The call must have completed and have a transcript.

Parameters

Name

Type

Required

Description

call_id

string

Required

ID of the call to analyze.

goal

string

Required

Overall purpose of the call. Gives the analysis context for interpreting the transcript.

questions

array

Required

Questions to answer about the call. Each question is a two-item array: the question text and the expected answer type, for example [["Did the customer agree to a follow-up?", "boolean"]].

bland_create_agent

Create agent

bland_create_batch

Batch create

bland_create_knowledge_base_from_text

Create knowledge base from text

bland_create_knowledge_base_from_web

Create knowledge base from web

bland_delete_block_rule

Delete block rule

bland_find_contact

Find contact

bland_get_account

Get account

bland_get_agent

Get agent

bland_get_batch_logs

Get batch logs

bland_get_contact

Get contact

bland_get_corrected_transcript

Get corrected transcript

bland_get_knowledge_base

Get knowledge base

bland_get_pathway

Get pathway

bland_get_persona

Get persona

bland_get_sms_conversation

Get sms conversation

bland_list_active_calls

List active calls

bland_list_agent_versions

List agent versions

bland_list_batches

List batches

bland_list_block_rules

List block rules

bland_list_contacts

List contacts

bland_list_knowledge_bases

List knowledge bases

bland_list_pathways

List pathways

bland_list_prompts

List prompts

bland_list_sms_conversations

List sms conversations

bland_list_tools

List tools

bland_send_call

Call send

Place an outbound AI phone call from Bland to one phone number, driven by a task prompt, a dashboard agent, a persona or a conversational pathway. Returns status, the new call_id and, when the call is part of a batch, the batch_id. The call runs asynchronously, so the result does not include a transcript. Use get_call with the returned call_id to read the outcome and transcript. Use create_batch to dial many numbers at once. Provide exactly one way to drive the call: task, agent_id, persona_id or pathway_id. Voice calls cost credits and reach a real phone.

Parameters

Name

Type

Required

Description

phone_number

string

Required

Phone number to call in E.164 format, with the plus sign and country code, for example +15551234567.

agent_id

string

Optional

UUID of a Bland agent in your organization to run the call. Use list_agents to find it. Requires an organization-level API key.

agent_version

string

Optional

Which saved version or environment of agent_id to use, for example production. Only valid together with agent_id.

dispositions

array

Optional

Outcome labels you define. After the call, the AI picks the best match from this list.

external_id

string

Optional

Your own identifier for the call, such as a CRM interaction ID. get_call accepts it in place of call_id.

first_sentence

string

Optional

Exact phrase the agent says first when the call connects.

from_number

string

Optional

Caller ID number to dial from, in E.164 format. Must be a number you own or imported from your own Twilio account. Omit to let Bland choose.

interruptibility

integer

Optional

How readily the agent stops talking when the caller speaks over it. Higher values yield more easily.

keywords

array

Optional

Words boosted in transcription, useful for names and jargon that are often misheard.

language

string

Optional

Language code that tunes transcription and speech for the call, for example babel-en or es.

max_duration

integer

Optional

Maximum call length in minutes. The call is ended automatically when this timer expires.

metadata

object

Optional

Custom metadata stored with the call and returned on call details and webhooks.

pathway_id

string

Optional

ID of a conversational pathway that scripts the call. Use list_pathways to find it. Replaces task.

pathway_version

integer

Optional

Pathway version to run: a version number, production or staging. Omit to use the production version.

persona_id

string

Optional

ID of a persona whose saved settings apply to the call. Use list_personas to find it.

persona_version

string

Optional

Persona version to use: production or draft. Only valid together with persona_id. One of: `production`, `draft`.

record

boolean

Optional

When true, the call audio is recorded and a recording_url appears on the call details.

request_data

object

Optional

Key-value data made available to the agent as variables in the prompt or pathway.

start_time

string

Optional

When to start the call, as an ISO 8601 date-time. If omitted or in the past, the call starts immediately.

summary_prompt

string

Optional

Custom instructions for how the post-call summary is written.

task

string

Optional

Instructions the AI agent follows during the call. Required when you do not pass agent_id, persona_id or pathway_id. Do not combine with pathway_id.

temperature

number

Optional

Randomness of the language model, from 0 for deterministic to 1 for creative.

timezone

string

Optional

IANA timezone for the call, for example America/New_York. Helps with scheduling and time-based answers.

transfer_phone_number

string

Optional

Phone number the agent can transfer the call to, for example when asked for a human, in E.164 format.

voice

string

Optional

Voice for the AI agent: a curated voice name such as Karen, or a voice ID from list_voices.

voicemail

object

Optional

Settings for how the agent behaves when it reaches voicemail, for example the message to leave or whether to hang up. Pass the object described in the Bland send call reference.

wait_for_greeting

boolean

Optional

When true, the agent waits for the other party to speak first instead of talking as soon as the call connects.

webhook

string

Optional

HTTPS URL that receives the full call details with transcript when the call ends.

webhook_events

array

Optional

Events to stream to the webhook while the call is in progress, for example call.

bland_stop_batch

Batch stop

bland_update_agent

Update agent

bland_update_inbound_number

Update inbound number

Update how an inbound phone number answers calls: its prompt or pathway, voice, webhook, transfer and call limits. Returns a status and message confirming the inbound number was updated. Use to reconfigure an inbound line. Use get_inbound_number to read its current settings first. The number must already be configured on your account; see list_inbound_numbers.

Parameters

Name

Type

Required

Description

phone_number

string

Required

Inbound phone number with country code and no plus sign, for example 15551234567. A US country code is assumed when none is given.

background_track

string

Optional

Background audio played during the call, for example office, cafe, restaurant or none.

block_interruptions

boolean

Optional

When true, the agent does not respond to interruptions from the caller.

fallback_number

string

Optional

Number that calls are forwarded to during Bland maintenance windows, in E.164 format.

first_sentence

string

Optional

Exact phrase the agent says first.

keywords

array

Optional

Words boosted in transcription, for names and jargon that are often misheard.

language

string

Optional

Language code for transcription and speech, for example en-US.

max_duration

integer

Optional

Maximum call length in minutes.

noise_cancellation

boolean

Optional

When true, background noise is filtered from the call audio.

pathway_id

string

Optional

ID of a conversational pathway for the number to follow. Overrides prompt.

pathway_version

integer

Optional

Pathway version number to use. Defaults to the production version.

prompt

string

Optional

Instructions for the inbound agent, with background information and the ideal conversation flow. Not needed when pathway_id is set.

request_data

object

Optional

Key-value data made available to the agent as variables during calls.

summary_prompt

string

Optional

Custom instructions for how the post-call summary is written.

timezone

string

Optional

IANA timezone for the number, for example America/New_York.

transfer_phone_number

string

Optional

Phone number the agent can transfer callers to, in E.164 format.

voice

string

Optional

Voice for the agent. Use a curated voice name or an ID from list_voices.

webhook

string

Optional

HTTPS URL that receives call details after each call to this number ends.

For more tools, view docs.

Build your Agent

Same auth pattern across LangChain, OpenAI, Anthropic, and Google ADK.

import { ScalekitClient } from "@scalekit-sdk/node";
import { createReactAgent } from "@langchain/langgraph/prebuilt";

const sk = new ScalekitClient(env.SCALEKIT_ENV_URL, env.SCALEKIT_CLIENT_ID, env.SCALEKIT_CLIENT_SECRET);

// Bland AI tools scoped to this user
const { tools } = await sk.tools.listScopedTools("user_123", {
  filter: { connectionNames: ["bland"], toolNames: [
    "bland_send_call",
    "bland_get_call",
    "bland_analyze_call"] },
  pageSize: 100,
});

const agent = createReactAgent({ llm, tools });
await agent.invoke({ messages: [{ role: "user", content: "Call the customer to confirm Thursday's appointment and summarize the outcome" }] });

import OpenAI from "openai";
import { ScalekitClient } from "@scalekit-sdk/node";

const sk = new ScalekitClient(env.SCALEKIT_ENV_URL, env.SCALEKIT_CLIENT_ID, env.SCALEKIT_CLIENT_SECRET);
const openai = new OpenAI();

const { tools } = await sk.tools.listScopedTools("user_123", {
  filter: { connectionNames: ["bland"] }, pageSize: 100,
});

const res = await openai.chat.completions.create({
  model: "gpt-5",
  messages: [{ role: "user", content: "Call the customer to confirm Thursday's appointment and summarize the outcome" }],
  tools,
});

// Execute the tool call with the user's vaulted Bland AI access token
await sk.tools.executeTool(res.choices[0].message.tool_calls[0], "user_123");

import Anthropic from "@anthropic-ai/sdk";
import { ScalekitClient } from "@scalekit-sdk/node";

const sk = new ScalekitClient(env.SCALEKIT_ENV_URL, env.SCALEKIT_CLIENT_ID, env.SCALEKIT_CLIENT_SECRET);
const anthropic = new Anthropic();

const { tools } = await sk.tools.listScopedTools("user_123", {
  filter: { connectionNames: ["bland"] }, pageSize: 100,
});

const msg = await anthropic.messages.create({
  model: "claude-sonnet-5",
  max_tokens: 1024,
  messages: [{ role: "user", content: "Call the customer to confirm Thursday's appointment and summarize the outcome" }],
  tools,
});

// Tool call runs with the user's vaulted Bland AI access token
await sk.tools.executeTool(msg.content, "user_123");

import { Agent } from "@google/adk/agents";
import { ScalekitClient } from "@scalekit-sdk/node";

const sk = new ScalekitClient(env.SCALEKIT_ENV_URL, env.SCALEKIT_CLIENT_ID, env.SCALEKIT_CLIENT_SECRET);

const { tools } = await sk.tools.listScopedTools("user_123", {
  filter: { connectionNames: ["bland"] }, pageSize: 100,
});

const agent = new Agent({
  name: "voice_ops_agent",
  model: "gemini-2.5-pro",
  instruction: "Bland AI tools scoped to this user",
  tools,
});

await agent.run("Call the customer to confirm Thursday's appointment and summarize the outcome");

Try these prompts

Copy any prompt into your agent. Each maps directly to a Bland AI tool. Click to copy, paste into your agent, done.

Calls

Copy the prompt

Copied

Call +1 415 555 0134 and confirm their appointment for Thursday at 3pm.

Copy the prompt

Copied

Show the transcript and summary of the last call to Dana Lee.

Copy the prompt

Copied

Which calls are in progress right now?

Batches and analysis

Copy the prompt

Copied

Start a batch of reminder calls to these 40 customers.

Copy the prompt

Copied

Stop the renewal batch that is still running.

Copy the prompt

Copied

Analyze yesterday's call and tell me whether the customer agreed to a demo.

Pathways and SMS

Copy the prompt

Copied

List our pathways and show the nodes in the booking flow.

Copy the prompt

Copied

Send an SMS to +1 415 555 0134 confirming tomorrow's appointment.

Copy the prompt

Copied

Add our pricing page to the knowledge base from its URL.

SEE HOW AUTH WORKS

Each user connects their own Bland AI access token once, and Scalekit sends it with every call. Access tokens stay vaulted, every call is scope checked, and every action is logged.

Authorize

Your user connects

Bland AI

once. We tie it to their identity and the meetings they approved — no shared bot account, no org-wide access

Who:

user ‘A’

when:

Once per user

access:

Limited to user

Store

Their

Bland AI

token lives in a vault scoped to them. User A's meetings are never reachable by an agent acting for user B, even on the same connection

vault:

encrypted

scope:

per-user

tokens:

auto-refreshed

Resolve

When your agent calls a

Bland AI

tool, we fetch the right token server-side. It never touches your agent, never appears in the LLM context, never shows up in your logs

speed:

~40ms

check:

before every call

seen by:

nobody

Audit

Every

Bland AI

tool call is logged — who triggered it, which meeting was fetched, what came back. 90 days of history, tied to the user who authorized it

history:

90 days

export:

SIEM-ready

logged:

every call

Test other agents

See the same per-user auth pattern across other AI and communication connectors.

GTM and RevOps Teams

Outbound prospecting agent

Searches Apollo for prospects matching your ICP, scores and ranks them, drafts personalized outreach in Gmail, and logs every send to Google Sheets. Mail goes out as the rep, not from a shared inbox.

Support and Ops Teams

Support ticket automation agent

Fetches new Zendesk tickets, drafts a reply from Notion knowledge base articles, digests what it cannot answer to Slack, and archives the rest, acting as the support agent rather than a shared API key.

Why Scalekit

Secure your agent's access. Connectors ship in minutes

01.

Shared tokens break per-user analytics

A shared Bland AI access token looks fine in a demo. In production every outbound call looks like one service account, and you cannot tell which user triggered it. Scalekit resolves the credential of the actual user who triggered the agent, never a shared bot.

// shared key
audit → bot_service_account

// scalekit
audit → user_abc ✓

02.

Authentication is not authorization

03.

Multi-tenancy is architectural

04.

Bland AI today. Ten connectors tomorrow.

“Our agents act across Salesforce, Gong, Google Drive, and more, on behalf of every customer. Scalekit behind the scenes meant we can keep adding tools without ever rebuilding how credentials or tool calling work.”

Venu Madhav Kattagoni

Head of Engineering / Von

FAQs

Frequently Asked Questions

Does the agent access Bland AI as the user or through a shared access token?

As the user. Each user connects their own Bland AI access token once, and Scalekit sends it with every call. Audit logs attribute every call, SMS, and pathway change to that user, not a shared service account.

Where is the Bland AI access token stored?

In Scalekit's managed AES-256 token vault, namespaced per tenant. To rotate it, the user reconnects with a new access token. Revocation is a single dashboard action. Credentials never appear in prompts, logs, or LLM context.

Can I limit what the agent does in Bland AI?

Yes. Filter by tool name in listScopedTools to expose only what you want, for example call history, transcripts, and analysis without bland_send_call or bland_delete_knowledge_base, which permanently deletes a knowledge base item. Scalekit also enforces scope checks before every API call.

What happens when a user revokes Bland AI access?

The connection is invalidated on the next tool call. Subsequent requests for that user fail closed with a clear error. Other users in the tenant remain unaffected. The event is logged for audit.

Can the agent end a call that is already in progress?

Yes, with bland_stop_call, which ends an in-progress call immediately. bland_list_active_calls shows calls that are queued or in progress, and bland_stop_batch halts the remaining calls in a running or scheduled batch.