Dapta
Dapta · Dapta: No-Code AI Voice, WhatsApp and SMS Agents That Answer, Qualify and Book Leads for Small Businesses, in English and Spanish
Dapta is a no-code platform for building AI agents that answer and place phone calls, reply on WhatsApp, SMS and website chat, qualify leads, book meetings and push the results into HubSpot, GoHighLevel, Salesforce or a spreadsheet. It suits small and mid-sized businesses, especially bilingual English and Spanish teams in the US and Latin America, that lose leads to missed calls and slow follow-up and want one credit-based tool instead of a voice platform plus a chatbot plus an automation service.
Best for
Small and mid-sized businesses in sales-led, appointment-driven fields such as home services, law, real estate, insurance and education, particularly bilingual English and Spanish teams in the US and Latin America, that want AI to answer every call and WhatsApp message, qualify leads and book meetings without hiring engineers
Not ideal for
Developers who want to pick and pay their own speech, model and telephony providers in code, teams that need local phone numbers in many countries without going through support, businesses that need predictable flat monthly pricing regardless of call minutes, and enterprises looking mainly for outbound B2B prospecting from a contact database
Who it's for
Owners, sales and admissions teams, and operations staff at small and mid-sized businesses, plus agencies that build AI phone and WhatsApp agents for them, with a strong base among Spanish-speaking businesses in the US and Latin America
Capabilities
- Voice Agents that answer inbound calls and run outbound calling Sequences from an uploaded CSV or Excel list or contacts a Flow Studio flow adds automatically, with calling windows, retries, batches of 1 to 500 contacts and a WhatsApp fallback for contacts who do not pick up
- Four ways to give an agent a line: buy a US number in Dapta (about 1,380 credits a month), verify an existing number for outbound caller ID, register your own phone as a test number (calls to it are free, on paid plans), or connect your own SIP trunk or PBX
- Call transfers in cold, warm or agentic warm mode, where a screening agent talks with the person who answers and decides whether to connect the caller, with business-hour rules and custom SIP headers
- Mid-call actions to check availability and book on Dapta Calendars or Cal.com, send a WhatsApp template, and end the call; voicemail detection that hangs up or leaves a tailored message
- Text Agents on WhatsApp Business (connected directly through Meta, with buttons, lists and carousels), a website chat widget, and SMS on US numbers after A2P 10DLC registration
- Brains (knowledge bases) built from up to 25 web pages, files or text sources per Brain, plus company context and dynamic variables in the prompt
- Post-call analysis on Voice Agents that extracts the fields you define and sends them to Google Sheets, HubSpot (find or create the contact, update the deal, add a note) or a Flow Studio flow
- CRM Pipeline Sync that lets a Text Agent move a contact's HubSpot deal or GoHighLevel opportunity to the matching stage after each conversation, or optionally open one for a contact who is already in the CRM
- Shared Inbox where teammates take over, add internal notes and hand conversations back, plus up to three automatic follow-ups per Text Agent
- Choice of language model per agent, including Dapta Auto (a model Dapta manages and updates) and named GPT and Gemini models, with ElevenLabs voice models for speech
- Flow Studio no-code automations started by a finished call, a webhook, a schedule or a form submission, with connections to HubSpot, Salesforce, Zoho, Gmail, Outlook, Google Sheets, Notion and Slack, OpenAI, Claude and Gemini nodes, and guided templates
- Dapti (beta), an in-app assistant that builds voice and text agents, flows and calling sequences from a chat, reports on calls against anonymized industry benchmarks, and asks for confirmation before changing anything live
- Dapta MCP server for claude.ai, Claude Desktop, Claude Code, ChatGPT, Codex and Cursor, with sign-in or API-key access and a preview-then-confirm step on every change
- AI Notetaker for Zoom, Google Meet, Microsoft Teams and in-person meetings (mobile app), priced per calendar rather than per user
Limitations
- Answered calls bill 333 credits ($0.333) per started minute, so a 45-second call costs a full minute, and calls that reach a personalized voicemail greeting bill as answered calls
- On WhatsApp lines connected directly through Meta, message delivery is charged by Meta to your own Meta account on top of Dapta's 40 credits per AI reply, and since October 1, 2026 Meta also bills service replies beyond 1,000 free per number each month; agents stop replying if the Meta account has no payment method
- Concurrency is capped by plan: Pro allows 5 live calls (2 inbound, 3 outbound), Scale-Up 20 and Prime 30, with extra slots at $10 a month each through support
- Phone numbers bought in-app are US-only; numbers in other countries are set up by support at quoted monthly prices (for example about $13 in Mexico, $28 in Colombia and $270 in Peru)
- US SMS requires A2P 10DLC registration with an EIN, one-time and monthly fees in credits, and a carrier review the docs put at about 10 to 15 days
- When the credit balance reaches zero, calls, messages and Dapti stop until credits are added (there is no overage), unless auto-recharge is turned on
- Remaining credits are not carried over if the subscription is cancelled or expires
- Dapti, the in-app assistant that builds agents and flows, is labeled beta, and on Free and Pro plans each Dapti answer is charged (about 150 credits on average)
- CRM Pipeline Sync supports only HubSpot and GoHighLevel, works only on V2 Text Agents, and never creates contacts; it acts only on people who already exist in the CRM
- Call recordings and transcripts are viewable by anyone with the call link unless the Public call link setting is turned off for that agent
- No self-hosted deployment of the agent platform is documented for self-serve plans; on-premise deployment is listed only for Enterprise
Use cases
- Answering a home-services or law-firm phone line after hours, qualifying the caller and booking a consultation on the team's calendar
- Calling every new Meta Ads or web-form lead within minutes from an outbound Sequence, retrying no-answers and falling back to WhatsApp
- Running admissions or sales conversations on a WhatsApp Business number in Spanish and English, with staff taking over from the shared Inbox
- Moving HubSpot deals or GoHighLevel opportunities to the right stage automatically after each WhatsApp conversation
- Reminding booked prospects by WhatsApp or SMS the day before an appointment to cut no-shows
- Asking Dapti or Claude through the MCP server which hours connect best and adjusting a calling sequence's schedule accordingly
Our take
Dapta's case rests on breadth for a specific buyer. A small business in the US or Latin America that sells by phone and WhatsApp, often in two languages, gets voice and messaging agents, outbound sequences, CRM sync and a meeting notetaker on one credit balance, with unusually complete documentation of what each call, message and number costs. That documentation also shows the tradeoffs clearly: billing by the started minute, Meta's separate WhatsApp delivery charges, two inbound calls at a time on Pro, and international numbers that go through support. Dapti and the MCP server lower the build effort, but Dapti is still beta and charged per answer below Scale-Up. Teams that mostly need an English-only AI receptionist or developer-level control over the voice stack may find simpler or cheaper options; bilingual, WhatsApp-heavy sales teams are where Dapta fits best.
Who should use it
Owners and sales, admissions or intake teams at small and mid-sized businesses that receive leads by phone, WhatsApp and web forms, want every one answered and qualified within minutes, and work in English and Spanish; agencies building AI agents for such clients; and teams already on HubSpot or GoHighLevel that want conversations to update the pipeline automatically.
Who should skip it
Developers who want to assemble and pay for their own speech, model and telephony providers, companies that need local numbers across many countries on day one, high-volume inbound operations that would outgrow the plan concurrency caps, and B2B teams looking for an outbound AI SDR that sources prospects from a contact database.
Strengths
- Phone, WhatsApp, SMS and website chat agents, outbound sequences, a CRM, automations and a meeting notetaker in one workspace and one credit balance
- Strong fit for bilingual English and Spanish businesses and for WhatsApp-first markets, including keeping the existing WhatsApp Business number and app
- Transparent, documented credit rates for every call outcome, message, number and assistant answer, with failed calls and test-number calls free
- Flexible telephony: buy a US number, verify your own, or connect a SIP trunk or PBX, with cold, warm and agentic warm transfers
- Dapti and the MCP server let non-technical owners or Claude and ChatGPT users build and adjust agents in plain language, with confirmation before live changes
- Results flow into HubSpot, GoHighLevel, Salesforce or Google Sheets without custom code
Weaknesses
- Calls bill by the started minute at about $0.33, and WhatsApp delivery on Meta-connected lines is billed separately by Meta
- Low concurrency on Pro (two inbound calls at once) can drop calls during a spike
- International numbers require a support request and can be expensive in some countries
- Many products and settings to learn compared with a single-purpose AI receptionist
- Dapti is still in beta and costs extra per answer below the Scale-Up plan
- No pay-as-you-go plan: beyond the one-time sign-up credits, ongoing use needs at least the $99 Pro subscription
Dapta pricing
Free
$0
- 7,000 one-time credits at sign-up
- Explore the platform and build agents
- 1 concurrent outbound call, no inbound
- Web calls 200 credits/minute; Dapti charged per answer
Pro
$99/month
Billed monthly
- 100,000 credits a month (about 300 call minutes)
- Unlimited agents and flows
- 5 users and 5 workspaces
- 5 concurrent calls (2 inbound, 3 outbound)
- Email and chat support, one tech session
- 7-day free trial with a card
Scale-Up
$499/month
Billed monthly
- 500,000 credits a month
- Unlimited users, 25 workspaces
- 20 concurrent calls
- Dapti included
- Fully assisted onboarding, priority chat and 4 one-on-one tech sessions a month
Prime
$1,499/month
Billed monthly
- 1.5 million credits a month
- Unlimited users, 70 workspaces
- 30 concurrent calls
- Dapti included
- Dedicated account manager and solutions engineer access
Enterprise
Custom
- Consumption-based credits
- Custom workspaces and storage
- Premium support with SLAs
- On-premise deployment option
Free tier limits: New accounts get 7,000 one-time credits to build and try agents; the free account has no monthly allowance or test numbers, allows one outbound call at a time and no inbound calls, and charges 200 credits a minute for browser web calls.
Note: 1,000 credits = $1. An answered call costs 333 credits per minute, rounded up to the whole minute; voicemail 75 credits; no answer or busy 45; failed calls and calls to your registered test number are free. An AI text reply costs 40 credits and a human Inbox message 10. A US number bought in Dapta costs about 1,380 credits a month; numbers in other countries are quoted by support. On WhatsApp lines connected directly through Meta, delivery is billed by Meta to your own Meta account. US SMS needs A2P 10DLC registration (one-time, vetting and monthly fees in credits). Dapti answers cost about 150 credits on average on Free and Pro; Flow Studio runs cost 1 credit, plus twice the model cost for AI nodes that use Dapta's connection. AI Notetaker's first calendar is free with 100 one-time recording hours; each extra calendar, or the first one after those hours run out, costs 30,000 credits a month. Extra concurrent calls are $10 a month each. Yearly billing saves 16.7%. Extra credits can be bought in $1 increments, with optional auto-recharge at a 20,000-credit balance; there are no overage charges, and plan credits do not carry over after cancellation.
Technical specs
Modalities
Voice, Text
API pricing
No separate API pricing is published; the Dapta MCP server is available to workspace owners and admins by sign-in or API key, and the actions it triggers use the workspace's credits
Where Dapta excels
After-hours intake for a law firm or home-services company
A Voice Agent on the business line answers every call, qualifies the request and books a consultation or job, with a warm transfer to staff during business hours.
Speed-to-lead for paid ads
A Flow Studio flow adds each new lead to an outbound Sequence, which calls it within the calling window, retries unanswered calls on the schedule you set and sends a WhatsApp template when the lead does not pick up.
Bilingual WhatsApp sales and admissions
A Text Agent answers on the existing WhatsApp Business number in the customer's language, sends catalogues as buttons or carousels, and hands the chat to a person in the shared Inbox when it matters.
Keeping a GoHighLevel or HubSpot pipeline current
CRM Pipeline Sync moves each existing contact's opportunity or deal to the right stage after a Text Agent conversation, so reps work from an up-to-date pipeline without manual updates.
Dapta vs. competitors
Dapta vs. Famulor
Famulor is a German no-code AI phone agent platform with booking, transfers, campaigns and WhatsApp, SMS and email follow-up, billed per second in euros from €0.15 a minute pay-as-you-go and hosted in Frankfurt. Dapta covers similar voice ground but centers on bilingual English and Spanish teams and WhatsApp-first selling in the Americas, bills US dollar credits per started minute on plans from $99 a month, and adds Text Agent CRM pipeline sync, a contact CRM and an AI notetaker.
Dapta vs. Harmony
Harmony is monday.com's voice AI for mid-market and enterprise revenue teams, priced at $0.39 a minute plus a $250 monthly platform fee and an implementation fee, with dispositions written to Salesforce, HubSpot, Zoho or monday.com. Dapta targets small businesses, with plans from $99 a month, no platform or implementation fee and about $0.33 an answered minute, and adds WhatsApp, SMS and web chat agents alongside voice.
Dapta vs. Vapi
Vapi is an API-first platform where developers choose and pay for their own speech, model and telephony providers and build agents in code. Dapta is no-code, bundles those components into one credit rate per minute, and adds WhatsApp agents, sequences, a CRM and automations for teams without engineers.
Frequently asked questions
What is Dapta?
Dapta is a no-code AI agent platform for small and mid-sized businesses. Its Voice Agents answer and make phone calls and its Text Agents reply on WhatsApp, SMS and a website widget, qualifying leads, booking meetings and sending results to tools such as HubSpot, GoHighLevel, Salesforce or Google Sheets. The same workspace includes a contact CRM, outbound Sequences, Flow Studio automations, an AI Notetaker, dForms and Dapti, an assistant that builds agents from a plain-language request.
How much does Dapta cost?
Dapta runs on credits, where 1,000 credits equal $1. Pro costs $99 a month with 100,000 credits, 5 users and 5 workspaces; Scale-Up costs $499 with 500,000 credits and unlimited users; Prime costs $1,499 with 1.5 million credits; Enterprise is priced by sales with consumption-based credits. Yearly billing saves 16.7%. An answered call uses 333 credits per started minute and an AI text reply 40 credits, so Pro covers roughly 300 call minutes a month. New accounts get 7,000 free credits at sign-up, and Pro has a 7-day trial that requires a card.
Does Dapta have a free plan?
A new account can explore the platform and build agents with 7,000 one-time sign-up credits. The free account does not include a monthly credit allowance or test numbers, web calls from the browser cost 200 credits a minute on it, and it allows one outbound call at a time and no inbound calls. Pro's 7-day trial needs a card and converts to a paid plan unless cancelled.
Which channels and languages does Dapta support?
Dapta agents work by phone (inbound and outbound), WhatsApp, SMS on US numbers and website chat. Dapta specializes in bilingual English and Spanish agents and says it supports many other languages; its WhatsApp page notes customers commonly run agents in English, Spanish and Portuguese.
Does Dapta work with HubSpot, GoHighLevel and Salesforce?
Yes. Post-call analysis can find or create a HubSpot contact, update its fields or deal and add a note with the summary and transcript. Text Agents can move HubSpot deals or GoHighLevel opportunities between pipeline stages for contacts already in the CRM. Salesforce connects through a Connected App for use in Flow Studio, and webhooks reach other systems.
Can I use Dapta from Claude or ChatGPT?
Yes. The Dapta MCP server at mcp.dapta.ai/mcp connects claude.ai, Claude Desktop, Claude Code, ChatGPT, Codex and Cursor to a Dapta workspace by signing in, or by API key for scripts. Workspace owners and admins can create agents, read call analytics, adjust sequences and edit flows; every change is shown as a preview and needs your confirmation.
Can I keep my existing phone number?
You can verify an existing number so outbound calls show it, or connect your own carrier through a SIP trunk so calls on your current numbers reach the agent. Numbers bought inside Dapta are US-only; numbers in other countries are arranged through support.
How many calls can Dapta handle at the same time?
Each plan has separate inbound and outbound limits: the free account allows 1 outbound call and no inbound calls, Pro 2 inbound and 3 outbound, Scale-Up 10 and 10, and Prime and Enterprise 15 and 15. Calls beyond the limit on either side do not connect until a live call ends. Support can add extra concurrent slots at $10 a month each.
Dapta vs Famulor: what is the difference?
Both are no-code AI phone agent platforms for small businesses with booking, transfers, campaigns and WhatsApp. Famulor is German, bills per second in euros from a pay-as-you-go balance and hosts data in Frankfurt. Dapta is built around bilingual English and Spanish teams and WhatsApp-first selling in the Americas, bills in US dollar credits per started minute, and adds a CRM, Text Agent pipeline sync and an AI notetaker in the same workspace.
Integrations & fit
Alternatives to consider
BEK Service GmbH
Small and mid-sized businesses and agencies, especially in Germany and the wider EU, that want a no-code AI phone agent able to answer, book and follow up on WhatsApp, with telephony and models included in one per-minute price
Freemium
monday.com
Mid-market and enterprise sales, insurance, finance and real estate teams in the US, UK, Australia or Singapore that call large lead volumes and want an AI SDR, dialer and CRM write-back in one vendor-managed product with SOC 2, ISO 27001 and HIPAA BAA options
PaidAbout Dapta
Dapta, founded in 2024 in Bogotá and also based in Phoenix, packages several pieces of a small company's lead-handling stack into one workspace. Voice Agents answer an inbound line or work through a contact list in outbound Sequences: they run on a US number bought in Dapta, a number you verify, or your own SIP trunk, detect voicemail, check availability and book on Dapta Calendars or Cal.com, send a WhatsApp template mid-call, and hand the caller to a person with a cold, warm or agentic warm transfer. Text Agents answer on a WhatsApp Business number (which your team can keep using in the WhatsApp Business app), on a website widget and, on US numbers, by SMS once the business completes A2P 10DLC carrier registration. Both kinds of agent answer from Brains built from your web pages and files and extract the fields you define after each conversation; Voice Agents can send those results straight to Google Sheets, HubSpot or a Flow Studio automation, and Text Agents can move a HubSpot deal or GoHighLevel opportunity to the right pipeline stage on their own. Around the agents sit a shared Inbox for human takeover, a contact CRM, Flow Studio for no-code automations, an AI Notetaker for Zoom, Meet and Teams, dForms for lead-capture forms, and Dapti, a beta assistant that builds agents, flows and calling sequences from a plain-language request and asks you to confirm before it changes anything live. Workspace owners and admins can also drive it from Claude, ChatGPT, Codex or Cursor through the Dapta MCP server. Everything runs on one credit balance (1,000 credits = $1): an answered call costs 333 credits a minute, rounded up to the whole minute, and an AI text reply 40 credits, with paid plans from $99 a month for 100,000 credits, which the docs say covers about 300 minutes of calls. What sets Dapta apart from European or developer-first voice platforms is its focus on bilingual English and Spanish agents and on WhatsApp alongside the phone. The tradeoffs: per-minute billing rounds up every started minute, WhatsApp delivery on lines connected through Meta is billed separately by Meta, the Pro plan allows only two inbound calls at once, numbers outside the US go through support (with high monthly fees in some countries), and the breadth of products means more to learn than a single-purpose AI receptionist.
Updates from Dapta
Since October 20, 2026, new Voice Agents use Dapta Auto, a language model Dapta selects for fast responses and upgrades over time, at the same per-minute rate. Existing agents keep their model until moved, and Dapta's docs say GPT-4o retires on November 9, 2026.
The AI Insider reported that Dapta closed a $6 million seed round led by Chingona Ventures, following a $1.4 million pre-seed in 2024, to expand its AI sales agent for small businesses.
From October 1, 2026, Meta bills the free-form replies an agent sends inside the 24-hour window, after 1,000 free service messages per number each month. Dapta's direct WhatsApp lines pass these charges to the customer's own Meta account; Dapta stops an agent's replies when the account has no payment method.
Are you the founder? Claim this listing →