Terms of Service

1. Acceptance of Terms

By accessing or using AgentCall ("the Service"), you agree to be bound by these Terms of Service. If you do not agree, do not use the Service.

2. Description of Service

AgentCall provides a programmable phone number API that enables AI agents to provision phone numbers, send and receive SMS messages, make and receive voice calls, initiate AI-powered voice calls where an AI handles the conversation autonomously based on a provided system prompt, configure phone numbers to be answered by an inbound AI voice agent that follows a user-supplied system prompt, opt in to call recording, receive post-call email summaries generated by AgentCall, optionally configure a pre-call context webhook so a customer-controlled endpoint can inject live context into the AI's system prompt for each inbound call, and receive event webhooks. The Service is also accessible via a hosted Model Context Protocol (MCP) server at api.agentcall.co/mcp and via a published Node.js SDK (npm install agentcall). Numbers are provisioned through licensed VoIP carriers; AgentCall does not guarantee that provisioned numbers will pass any third-party platform's carrier-type or verification checks.

Outbound coverage. Outbound SMS and voice from AgentCall numbers are supported to United States and Canada destinations only at this time. Outbound to other countries is not supported and may return a carrier error. Inbound calls and SMS to AgentCall numbers may originate from anywhere. International outbound coverage may be added in the future and will be announced when available.

Authentication. Users may sign in with an email magic link or with a GitHub OAuth account.

3. Account and API Keys

You are responsible for maintaining the confidentiality of your API keys and for all activity that occurs under your account. You must notify us immediately of any unauthorized use.

4. Acceptable Use

You agree not to use the Service to:

• Send unsolicited messages (spam)
• Engage in fraudulent or illegal activity
• Violate any applicable laws or regulations
• Interfere with or disrupt the Service
• Use AI voice calls for impersonation, deception, or automated robocalling without recipient consent
• Resell access without authorization

You are solely responsible for complying with all applicable call-recording, call-monitoring, automated-dialing, and AI-disclosure laws in every jurisdiction where you, your callers, or your recipients are located. This includes obtaining any consent or providing any notice required by federal, state, provincial, or international law before recording, transcribing, structuring, or otherwise processing a call. Where local law requires you to disclose that a call is being handled by AI, that it may be recorded, or that it may be processed to generate memory or summary data, providing that disclosure to your callers is your responsibility.

5. Billing and Payment

Free plan. Users on the Free plan are not charged for usage. The Free plan currently includes 1 phone number, 10 SMS per month, 5 voice minutes per month, 5 inbound AI voice minutes per month (the "Inbound AI Trial"), and 5 verification code extractions per month. The Inbound AI Trial requires no payment method on file and resets on the first day of each calendar month (UTC). Free plan accounts do not have a metered Stripe subscription attached and AgentCall does not invoice Free plan accounts for any per-message, per-minute, or per-number usage. Free plan limits and trial allowances may change with notice.

Pro plan and metered usage. The Pro plan is billed monthly at the rate published on agentcall.co/#pricing, plus usage-based fees for SMS, voice minutes, AI voice call minutes, call recording, and per-number subscriptions. A valid payment method is required to enable any feature that bills usage past Free plan allowances, including inbound AI voice past the Inbound AI Trial. You agree to pay all fees associated with your use of the Service. Failure to pay may result in suspension or termination of your account.

AI voice billing tiers. AI voice calls (inbound and outbound) are billed under one of two tiers, configurable per phone number:

• Managed at $0.40/minute (default). AgentCall handles the underlying language model API and bills you a single invoice that covers the model, carrier, audio, Memory extraction, and infrastructure costs.
• Bring Your Own Key (BYOK) at $0.10/minute. You provide an OpenAI API key that AgentCall uses to make model calls on your behalf for AI voice calls on numbers configured for BYOK. You are billed by OpenAI directly for the model usage under your own OpenAI account, and AgentCall bills you the $0.10/minute service fee that covers carrier, audio, Memory extraction, and infrastructure. BYOK requires the Pro plan and is currently OpenAI-only. You may switch any number between Managed and BYOK at any time from your dashboard. Switching back to Managed disables further use of your stored OpenAI key for that number.

BYOK responsibilities. When you enable BYOK on a number, you are responsible for maintaining a working OpenAI API key with sufficient quota, for any charges OpenAI bills you directly, and for compliance with OpenAI's own terms of service. AgentCall is not responsible for OpenAI rate limits, outages, content policy decisions, or billing on the customer's OpenAI account. If your provided OpenAI key fails (invalid, revoked, out of quota, or content-filtered), AI voice calls on numbers configured for BYOK may fall back to a hangup or to Managed billing at AgentCall's discretion.

Trial exhaustion. Once the monthly Inbound AI Trial is used up, additional incoming calls to numbers configured for AI voice are answered with a hangup and no charge is incurred. The trial counter resets on the first day of the next calendar month (UTC). There is no pay-as-you-go overage path on the Free plan; to use inbound AI voice past the trial within the same month, you must upgrade to the Pro plan.

No refunds for usage already incurred. Cancellation stops future charges; minutes, messages, and recordings already consumed in the current billing period are non-refundable.

6. Limitation of Liability

The Service is provided "as is" without warranties of any kind. AgentCall shall not be liable for any indirect, incidental, or consequential damages arising from your use of the Service.

In no event shall AgentCall's total aggregate liability exceed the amounts paid by you to AgentCall in the twelve (12) months preceding the claim.

7. Termination

We may suspend or terminate your access at any time for violation of these terms. You may terminate your account at any time by contacting us at contact@agentcall.co.

8. Changes to Terms

We reserve the right to modify these terms at any time. Continued use of the Service after changes constitutes acceptance of the updated terms.

9. Telecommunications Compliance

Users must comply with all applicable telecommunications regulations, including but not limited to TCPA, CAN-SPAM, STIR/SHAKEN, and any other federal, state, or international telecom laws.

• You must obtain proper consent before sending automated messages or making automated calls
• You must not use phone numbers for verification fraud, account farming, or spam
• AgentCall reserves the right to immediately suspend accounts that violate telecommunications regulations

SMS Opt-Out Handling: AgentCall automatically processes opt-out keywords (STOP, UNSUBSCRIBE, CANCEL, END, QUIT) received via SMS. When a recipient opts out, they are blocked from receiving further messages from your account and receive a confirmation reply. Recipients may opt back in by replying START. You must not attempt to circumvent this opt-out mechanism, including by sending messages from a different number to opted-out recipients. Violations may result in immediate account suspension.

Call Recording and Two-Party Consent: Call recording is opt-in and disabled by default. When you enable recording on an AI voice path (inbound or outbound), AgentCall automatically prepends a recording disclosure to the spoken first message in two-party-consent jurisdictions (CA, FL, IL, MD, MA, PA, WA, NV, NH, MT, CT, DE) unless your configured first message already references recording. For programmatic outbound calls (POST /v1/calls/initiate), where you control the audio script directly, disclosure is your responsibility. You are solely responsible for compliance with consent and recording laws applicable to your use case and jurisdiction.

Post-Call Summary Emails: When you configure a notification email address on an inbound AI number, AgentCall sends a plain-English summary of each call (including caller phone number, intent, urgency, and a 1-2 sentence ask) to the address you specify. You are responsible for the addresses you configure and for ensuring the recipient is authorized to receive call data. AgentCall does not send these summaries to addresses you have not configured.

Pre-Call Context Webhooks: If you configure a pre-call context webhook URL on a phone number, AgentCall will POST call metadata (caller phone number in E.164 format, the internal number ID, an opaque call ID, and a timestamp) to your URL on every inbound call connect. The request is HMAC-signed with the shared secret you provide. You are solely responsible for the security, uptime, and legal compliance of any webhook endpoint you configure, including PCI-DSS (if processing payment data), HIPAA (if handling protected health information), GDPR (if handling EU resident data), and any other regulatory regime applicable to your use case. Webhook timeouts and errors do not prevent the call from connecting (fail-open behavior); AgentCall will proceed with the static system prompt if your endpoint does not respond or returns an error.

Call Reports and Cross-Call Memory: When the Cross-Call Memory feature is enabled on your account, AgentCall extracts structured Call Reports and Memory entries from the transcripts of completed AI voice calls on your numbers. Call Reports may include the caller's phone number, stated name, intent, urgency, mentioned entities, facts, decisions, commitments, tasks, preferences, unresolved items, and an owner brief. Each Memory entry preserves a verbatim quote from the transcript line that produced it so you can audit accuracy. Memory is scoped to your account and is not visible to other AgentCall customers. The feature is on by default and can be turned off at any time from your dashboard; turning it off stops extraction, stops new memory writes, and stops memory reads on subsequent calls, effective immediately. You can view, edit, expire, and delete Memory entries and Call Reports from your dashboard. You are solely responsible for any notices or consents required under applicable law before processing callers' speech into structured memory, including any jurisdiction-specific obligation to inform callers that their conversation may be transcribed and processed to generate persistent records. AgentCall acts as a processor for caller data on your behalf; you are the controller of that data.

10. Intellectual Property

AgentCall retains all intellectual property rights to the Service, including its API, documentation, brand, and all related materials. Users are granted a limited, non-exclusive, revocable license to use the API solely for the purpose of integrating with the Service.

You may not reverse engineer, decompile, disassemble, or create derivative works of the Service or any part thereof.

Open-Source Templates: AgentCall publishes reference implementations and integration templates as standalone open-source projects under their respective licenses, including the agentcall-hermes-bridge Cloudflare Worker template at github.com/Kintupercy/agentcall-hermes-bridge (MIT License). These templates are provided as-is for reference and integration purposes, are not part of the AgentCall Service, and are governed by their own license terms rather than this Agreement's license grant. AgentCall makes no warranty about their fitness for any particular use.

11. Indemnification

You agree to indemnify, defend, and hold harmless AgentCall, its officers, directors, employees, and agents from and against any claims, liabilities, damages, losses, and expenses (including reasonable legal fees) arising out of or in any way connected with your use of the Service, your violation of these Terms, or your violation of any applicable laws or regulations.

12. API Usage and Fair Use

AgentCall reserves the right to rate limit, throttle, or suspend access for abusive or excessive API usage. Service level objectives are provided on a best-effort basis and do not constitute a guaranteed service level agreement unless separately agreed upon in writing.

13. Governing Law

These Terms shall be governed by and construed in accordance with the laws of the State of Delaware, United States, without regard to its conflict of law provisions. Any disputes arising under or in connection with these Terms shall be resolved exclusively in the state or federal courts located in Delaware.

14. Privacy

Your use of the Service is also governed by our Privacy Policy.