Autocalls
Also known as: autocalls.ai, Autocalls.ai, MULTICODE SRL
Autocalls is a no code platform for AI voice agents that place and answer calls, run WhatsApp, SMS and chat conversations, book meetings and look up data mid call through custom API tools and remote MCP servers, in more than 100 languages. Teams pick the conversation mode, transcriber and model, from GPT-5 to Gemini, design flows visually, run campaigns from lead lists and automation triggers, choose all EU or all US routing, and manage everything through a REST API or an OAuth MCP connector for Claude and ChatGPT.
Autocalls is an AI voice agent platform from MULTICODE SRL, sold directly and through a white label program to agencies, with alternatives pages aimed at Vapi, Retell, Bland and Synthflow users and a migration path for agencies that lost Synthflow's in product reselling in September 2026. It covers phone agents for reception, answering services, appointment booking, cold calling, call centers and collections, plus WhatsApp and web chat agents, with the AI model, voice and transcription included in the per minute price and no API keys to manage.
An assistant is set up in a guided overview or an advanced builder with version history, choosing a mode (Dualplex in beta, speech to speech, or a pipeline with a separate transcriber), a model such as GPT-5 Mini, GPT-5 Realtime, GPT-4o or Gemini Flash, a transcriber such as Deepgram, and a voice, including cloned voices. A drag and drop flow builder designs conversation paths, knowledge bases hold FAQs and documents, and built in tools handle actions such as transfers and booking.
Custom mid call tools call any endpoint with any standard HTTP method, headers, parameters and a 1 to 30 second timeout, and remote MCP servers add external tools to voice and chat conversations. After a call, AI extracts defined variables and sends them to a webhook. Campaigns reach lead lists by call, WhatsApp or SMS with parallel dialing, and a built in automation platform with triggers, actions, versioning and ready made templates starts calls from form submissions, Google Sheets rows and other events, or looks up orders and reschedules Cal.com appointments mid call. Conversations can be handed to a person by switching AI off for that thread.
Plans cost $34 a month with 120 minutes and $0.20 an extra minute, Pro $129 with 700 minutes at $0.16, Agency $249 with 1,700 minutes, unlimited assistants and 500 parallel calls at $0.09, and Agency White Label $419 with 3,500 minutes, client sub accounts, custom domains and Stripe billing, with two months free on yearly billing and no setup fee or revenue share.
Autocalls holds ISO 27001:2022 (registration I250699/01/EN, valid to December 2028) and ISO 9001, offers HIPAA with a BAA as an enterprise option and a HIPAA mode per sub account, encrypts with AES 256 and TLS 1.3, and lets customers set data retention for calls, leads, conversations and SMS. Customers can route everything through the EU or through the US, with OpenAI, ElevenLabs, Deepgram and Cartesia as AI subprocessors.
Autocalls fits agencies and resellers building voice agents for many clients, and SMBs that want an all inclusive per minute price with real model choice, API tools and regional routing. It fits less well for buyers who need a SOC 2 report specifically, approval steps before an agent acts, or memory of earlier conversations with the same caller.
Vendor details
Canonical URL
https://autocalls.ai
Category
Voice agent
Subcategory
All inclusive voice and messaging agent platform with model choice, custom mid call API tools, MCP servers, campaigns, an automation engine, EU or US routing and white label resale
Funding status
MULTICODE SRL, operating Autocalls.
Company status
independent
Use cases & customers
Primary use cases
Target customers
Deployment options
Integrations
Custom mid call tools call any HTTP endpoint with any standard method, headers and parameters; remote MCP servers add external tools to voice and chat. Named integrations include HubSpot, GoHighLevel, Zoho CRM, Cal.com, Google Sheets and calendars, with a built in automation platform of triggers, actions and templates, webhooks after calls, WhatsApp Business, SMS and SIP trunks. A REST API covers assistants, calls, campaigns, leads, conversations, knowledge and automations, and an OAuth MCP connector lets Claude and ChatGPT build assistants and run campaigns. Models from OpenAI and Google, with ElevenLabs, Deepgram and Cartesia for speech.
In practice
A dental agency builds a receptionist on the Agency plan, adds a Cal.com tool to cancel and reschedule visits mid call, and routes all data through the EU for its German clinics.
An ecommerce brand adds a custom mid call tool that calls its order API with the caller's order number, so the agent reads back delivery status and sends a WhatsApp follow up from the same platform.
A real estate agency triggers an outbound call whenever a lead submits a web form, and post call variables on budget and timing post to a webhook that updates its CRM.
Sources & related URLs
Related / legacy domains
Research sources
Agentic Index coverage score
10.5 / 14 capabilities · 75%
| Integrations & Tool Calling | Full |
|---|---|
|
Custom mid call tools let an assistant call the business's own API during a conversation: the admin sets the endpoint, any standard HTTP method for reading, creating, updating or deleting records, headers, JSON or form encoded body parameters and a 1 to 30 second timeout, with no code. Remote MCP servers add external tools to voice and chat conversations, built in tools cover transfers and booking, and tutorials show order lookups from Google Sheets and Cal.com cancellations and reschedules mid call, with HubSpot, GoHighLevel and Zoho CRM integrations alongside. Sourcedocs.autocalls.ai/ai-assistants/custom-toolsread 2026-10-18 |
|
| Workflow Orchestration | Full |
|
A drag and drop flow builder designs conversation paths visually, and a built in automation platform runs flows made of a trigger and a series of actions, passing data from earlier steps into later ones, with publishing, version history, rollback and documented execution limits. Flows can start calls, look up data during a call and push results to calendars, sheets, CRMs and notification tools, and ready made automation templates can be applied and proven with a test run. Sourcedocs.autocalls.ai/automation-platform/building-flowsread 2026-10-18 |
|
| Knowledge Grounding & RAG | Full |
|
Knowledge bases give assistants access to FAQs, documents and other sources, are attached in an assistant's in call settings alongside its tools, variables and system prompt, and can be created and updated through the API and the MCP connector. Knowledge bases also move between accounts in white label data migration, so a reseller can reuse them across client sub accounts. Sourcedocs.autocalls.ai/conversation-design/knowledge-basesread 2026-10-18 |
|
| Human Oversight & Guardrails | Partial |
|
A transfer tool hands live calls to a person, and for WhatsApp and chat the API can disable AI replies on a single conversation so a human agent takes over, then enable them again. Version history on assistants and automations lets a team see what changed and roll a setting back. Tools and campaigns run without a person approving an action before it happens. Sourcedocs.autocalls.ai/api-reference/conversations/disable-airead 2026-10-18 |
|
| Security, Identity & Governance | Full |
|
MULTICODE SRL holds ISO/IEC 27001:2022 under registration I250699/01/EN, issued December 2025 and valid to December 2028 for the whole platform, plus ISO 9001, and offers HIPAA with a BAA as an enterprise option and a HIPAA mode per white label sub account. Role based access, multifactor authentication and logged access events apply across systems, white label clients get their own logins with user limits, and customers set retention periods for calls, leads, conversations and SMS. No SOC 2 report is claimed. Sourceautocalls.ai/securityread 2026-10-18 |
|
| Observability & Auditability | Full |
|
Inbound call insights show each call's transcript, recording and the automations it triggered, conversations can be filtered and reviewed, and through the MCP connector a team can list calls, read full transcripts and post call evaluations, then fix the assistant or its tools. Version history records what changed on each assistant and flow, and the security program logs every access event for audit. Sourcedocs.autocalls.ai/inbound-calls/insightsread 2026-10-18 |
|
| Memory & State Persistence | Not documented |
|
Leads carry custom variables and secondary contacts into each call, conversation history is kept for review, and after call extraction sends variables to webhooks, while retention settings decide how long calls, leads and conversations are stored. None of these is read back by the assistant as memory of a caller on the next conversation; each call starts from the prompt, variables and knowledge base. Sourcedocs.autocalls.ai/leads/indexread 2026-10-18 |
|
| Deployment & Data Residency | Full |
|
Customers choose 100% EU routing, with all voice processing, transcription and storage inside the EU and EU routed voice processing through an ElevenLabs Enterprise partnership, or 100% US routing, with all data processed and stored in the United States. The white label edition runs on the partner's own domain with separate client sub accounts, and phone numbers can come in over SIP trunks from carriers such as DIDLogic. Sourceautocalls.ai/securityread 2026-10-18 |
|
| Prebuilt Agents / Templates / Packs | Partial |
|
Ready made automation templates, listed with the parameters each needs, create automations that are then proven with a test run, and white label partners publish reusable prompt templates for their own users. Solution pages describe receptionists, answering services, booking and cold calling agents, but assistants are assembled from a prompt, a voice, a model and tools rather than adopted from a catalog of finished agents. Sourcedocs.autocalls.ai/white-label/prompt-templatesread 2026-10-18 |
|
| Triggers & Channel Coverage | Full |
|
Assistants answer inbound calls, run outbound campaigns over lead lists with caller ID control and parallel dialing up to 500 or 1,000 calls, and send call, WhatsApp and SMS campaigns. Automation triggers start calls from events such as a form submission or a new row in a live Google Sheet, WhatsApp numbers take calls and messages, chat widgets sit on websites, and the REST API and the Claude and ChatGPT connector can add leads and start campaigns, in more than 100 languages. Sourcedocs.autocalls.ai/campaigns/indexread 2026-10-18 |
|
| Model Flexibility & Routing | Full |
|
Each assistant picks a conversation mode, Dualplex in beta, speech to speech or a pipeline, then a transcriber such as Deepgram for pipeline mode, and a language model such as GPT-5 Mini, GPT-5 Realtime, GPT-4o or Gemini Flash 2.0 and 2.5, with guidance on which pairings suit fast turn taking or long replies. Pricing counts more than 40 models from seven providers, all included in the per minute price, and voices include cloned ones. Sourcedocs.autocalls.ai/ai-assistants/assistant-configurationread 2026-10-18 |
|
| APIs / SDKs / MCP Extensibility | Full |
|
The REST API, with authentication and a reference covering the current user, assistants (list, create, update, delete), calls, campaigns, leads, conversations including AI takeover controls, and automations with templates and rollback, lets outside code build and run agents. A remote MCP server secured with OAuth 2.0 connects Claude and ChatGPT to the account to create assistants, edit knowledge bases and mid call tools, run campaigns, add leads and read transcripts. Sourcedocs.autocalls.ai/api-reference/introductionread 2026-10-18 |
|
| Testing, Debugging & Optimization | Partial |
|
Every assistant has a test panel with three tests, a phone call, a browser call and a text chat for fast iteration, each running against the saved version, with a separate guide for hearing inbound assistants. The advanced builder adds an advisor, and automation templates are proven with a real test run before use. There are no scored test cases or simulated caller suites that grade an assistant before release. Sourcedocs.autocalls.ai/ai-assistants/testingread 2026-10-18 |
|
| Browser / Computer-use | Not documented |
|
Assistants act through phone and WhatsApp calls, chat, custom API tools, MCP servers and the automation platform's integrations with sheets, calendars and CRMs. Nothing has an assistant open a browser or desktop session to operate another application's screens, and the browser call option in the test panel lets a team talk to its own assistant rather than giving the agent a browser. Sourcedocs.autocalls.ai/ai-assistants/tools-and-functionsread 2026-10-18 |
|
The Agentic Index coverage score grades every vendor Full, Partial or Not documented against the same 14 buyer facing capabilities, from public evidence only. Each capability links to how all vendors in the index score on it. How this evidence is graded
Pricing
Starter $34 a month (120 min); Pro $129 (700 min); Agency $249 (1,700 min); Agency White Label $419 (3,500 min)
Monthly plan with included minutes, then a per minute rate that falls by tier, with model, voice and transcription included
Included quota
Starter 120 minutes, Pro 700, Agency 1,700 with 500 parallel calls, Agency White Label 3,500 with 1,000 parallel calls.
What is public
All four plan prices, included minutes, per minute overage rates and parallel call limits.
Billing mechanics
Monthly or yearly plans with included minutes and per minute overage, all inclusive of model and voice costs.
Cost watchouts
WhatsApp, SMS and chat replies carry their own rates, phone numbers or SIP trunks cost extra, and HIPAA is an enterprise option.
Variable cost rationale
Minutes beyond the bundle bill per minute, from $0.20 on Starter down to $0.09 on Agency plans, so high volume teams move up a tier to cut the rate.
Additional watchouts
Starter's $0.20 overage is more than double the Agency rate.
Overage / add-ons
Extra minutes bill at $0.20 on Starter, $0.16 on Pro and $0.09 on both Agency plans; WhatsApp, SMS and AI chat replies are priced separately.
Sales call required
No, self serve available
Free / trial
No free tier stated; yearly billing gives two months free
Lowest paid plan
Starter at $34 a month with 120 minutes, then $0.20 a minute
Commercial notes
No revenue share or per client fee for resellers; yearly billing gives two months free.
Key ambiguities
No free trial is stated on the pricing page, and enterprise and HIPAA pricing are quoted.
Cancellation / refund
Every plan can be cancelled monthly; no setup fee or revenue share.
Support SLA / resale
Agency White Label adds the partner's own brand, domain, client sub accounts, pricing plans and Stripe billing.
Missing data
Enterprise and HIPAA pricing, messaging rates on direct plans.
Related vendors
- AgentVoice — AgentVoice builds and runs inbound and outbound AI voice agents, and…
- aiola — Voice agents for field teams on Salesforce
- Aircall — Aircall's AI Voice Agents answer and place calls on Aircall's own…
- Allo — Allo is a business phone system for small teams with an AI…
- AviaryAI — Outbound AI voice agents for credit unions, banks and insurers,…
- Avoca — AI front office for home services