SadaSend is the transactional email and deliverability operating system with built-in safety guardrails, engineered for modern developers and autonomous AI agents to prevent runaway loops, staging leaks, and surprise invoices.
Give your agent email.
Keep your peace of mind.
The transactional email API with built-in safety guardrails. Scoped keys, recipient allowlists, 1-click Slack approvals, and real-time bounce shields—so autonomous agents and web apps never email dead mailboxes or leak staging data.
6,000 emails free · no card · every guardrail on every plan
// The safety rules, set once when the key is created.
// The agent gets the key. It cannot widen any of this.
{
scopes: ['send', 'read', 'domains'], // all 12 safe tools
allowlist: ['@yourteam.com'], // who it may reach
rateLimitMax: 50, // how fast
rateLimitWindow: '1h',
mode: 'approval', // who signs off
}Deliverability protected by direct mail delivery architecture and pre-send checks
Major mailbox providers (Gmail, Yahoo, Apple, Microsoft) immediately drop non-compliant traffic. SadaSend delivers directly to destination mail servers, cryptographically aligns SPF/DKIM/DMARC, injects one-click unsubscribes, and filters invalid recipients before any email hits the wire.
- SPF · DKIM · DMARC
- Cryptographically verified before your first production email is sent
- RFC 8058 Unsubscribe
- Machine-readable one-click unsubscribe headers injected automatically
- Complaint Guard
- Monitored across a rolling 7-day window (evaluated every 5 min): warning at 0.08%, throttling at 0.10%, suspension at 0.25%
- Instant Suppression
- Hard bounces and complaints automatically suppressed account-wide
Everything modern apps and AI agents need to send, receive, and protect email
Built from the ground up for developer peace of mind. Whether you are running autonomous AI agents, transactional SaaS loops, or contact forms, SadaSend replaces fragile glue code with native infrastructure.
Autonomous AI Agent Safety Stack
Loop circuit breakers, spend caps, and human-in-the-loop Slack approvals.
Handing an autonomous agent an unbounded email API key is a serious risk. SadaSend gives agents strictly scoped keys with hourly velocity limits, a default $0 overage ceiling, and an approval queue that holds unverified sends until a human taps Approve in Slack or the web console.
- 12 hosted MCP tools for Claude Desktop, Cursor, LangChain, and CrewAI
- Hard hourly send limits and automatic loop circuit breakers
- Interactive Slack notifications with 1-click human approval holds
Zero-Backend Contact Forms API
Point any HTML form to SadaSend. Get instant inbox alerts and lead spreadsheets.
Turn any static landing page, HTML form, or React component into an automated lead capture engine. Submit directly to our public forms endpoint with a public token. Automated bots are silently trapped with invisible honeypots without ugly captchas, and submissions populate your dashboard lead grid.
- Embed in HTML or React in 60 seconds with sada_form_... public keys
- Instant email notifications and custom branded auto-responders
- Searchable leads spreadsheet with instant CSV export and webhook dispatches
Inbound MX Routing & Token Optimizer
Process inbound emails with up to 85% LLM token context reduction.
Receive incoming emails directly on your domain. Our parser strips out noisy MIME headers, redundant signatures, and tracking pixels, converting multi-part emails into clean Markdown. Pass clean content directly into RAG pipelines, OpenAI, Anthropic, or Gemini without wasting costly context tokens.
- Direct custom domain inbound routing via MX records to inbound.sadasend.com
- Noise and signature stripping for ~85% LLM context token savings
- Cryptographic HMAC-SHA256 webhook deliveries with dead-letter queue retries
DMARC Forensic Radar (RFC 7489)
Ingest aggregate reports, uncover spoofing, and achieve p=reject without third-party fees.
Never pay a third-party service $50 to $200 every month just to decode DMARC XML reports. SadaSend automatically ingests and decodes aggregate reports from Google, Microsoft, Yahoo, and Apple, providing a live health scorecard and threat radar to safely reach p=reject.
- Automatic rua=mailto: ingestion for Google, Microsoft, Yahoo, and Apple reports
- Real-time alignment scorecard tracking SPF, DKIM, and spoofing attempts
- Included across all plans with zero third-party subscription overhead
In-Flight Bounce Shield & Studio
Block invalid recipients in-flight before emails hit the wire. Zero quota debit.
High bounce rates damage sender reputation and trigger mailbox provider blocks. SadaSend validates recipient MX records, syntax, and disposable mailboxes in real time. Blocked bad addresses are refused before send and never debit your monthly quota, preserving both reputation and billing.
- In-flight syntax, MX existence, and disposable mailbox verification
- Pre-send refusal guarantee: bad addresses never count against plan quota
- Interactive single and bulk CSV email verification studio in the dashboard
1-Click Domain Connect & SPF Merger
Instant DNS records with smart SPF conflict resolution preventing 10-lookup errors.
Setting up DNS authentication is traditionally error-prone. SadaSend provides 1-click Domain Connect for major DNS providers (Cloudflare, Namecheap, GoDaddy, Route 53) and features a smart SPF auto-merger that prevents breaking the strict RFC 7208 10-DNS-lookup limit.
- 1-Click Domain Connect and guided verification for 7 major DNS providers
- RFC 7208 SPF conflict auto-merger prevents RFC 10-DNS-lookup failures
- Automatic 2048-bit DKIM key generation with sada selector isolation
Start with 6,000 free emails every month
No credit card required. Upgrade when you need more volume: Starter ($12/mo), Pro ($19/mo), or Scale ($79/mo) with annual savings. Transparent overage: $0.60, $0.50, or $0.40 per 1k emails. Zero per-seat charges, ever.
Traditional email APIs grant all-or-nothing access. SadaSend gives you built-in safety controls.
Handing an AI agent an ordinary API key lets it email anyone in the world if it loops or hallucinates. SadaSend provides built-in safety controls—recipient allowlists, hourly send caps, and 1-click approval holds—included on every plan.
keys.createEnforced at the API before the send reaches the queue — not in a doc, not in a lint rule, and not dependent on the agent behaving.
Every safety gate can be configured per API key via the developer dashboard or REST API.
Read security architecture & safety controlsSimulate Runaway Agents & Staging Leaks
Explore simulated server-enforced safety guardrails right in your browser. Watch how velocity tripwires and allowlists cut off hallucinated dispatches before an email can leak.
Simulate Agent Infinite Hallucination Loop
An LLM agent gets stuck in a recursive loop and attempts to send dozens of emails in rapid succession.
0 / 5 max
0
Click "Trigger Loop" to simulate a runaway agent send burst.
Observes exact status codes and error responses.
The surfaces where you approve, inspect, and debug
Automated email should never be a black box. Inspect the real dashboard interfaces where your team clears autonomous agent drafts, audits DKIM alignment, reviews DMARC aggregate forensics, and manages contact form leads.

Every held message can be approved or rejected in the dashboard, from Slack, or through the approvals API.
Explore architecture documentationHow SadaSend prevents runaway email
The barrier sits in front of the queue, not in a prompt or system instructions. Every send is checked by server rules before anything is dispatched.
Agent initiates tool call
Your agent (Claude, Cursor, LangChain) invokes send_email with recipient and body. It cannot alter the key’s preset limits.
SadaSend evaluates policy gate
Every send is verified against the key’s server-enforced scopes, recipient allowlist, hourly rate limits, and suppression history.
Instant dispatch or human sign-off
Allowlisted transactional sends are accepted and queued immediately. Out-of-allowlist sends are blocked; keys configured in approval mode hold dispatches in pending_approval for 1-click human clearance.
- Zero LLM Training
- Customer email content, variables, and payloads are never used to train or fine-tune models.
- Server-Enforced Key Scopes
- Agents hold keys that can only send or read — never mint keys, delete domains, or remove suppressions.
- Cryptographic Webhooks
- HMAC-SHA256 signatures with timestamp replay protection on all delivery events.
- Append-Only Audit Log
- Every attempt, blocked send, and delivery event is recorded with millisecond precision for audit inspection.
Want to see how the API enforces rate limits, token scopes, and allowlists?
Read full architecture specificationsUnder ten lines to a delivered email
Install, set a key, send. Zero-dependency official TypeScript and Python SDKs live on npm and PyPI, alongside our hosted Model Context Protocol (MCP) server for AI agents.
https://mcp.sadasend.com// Universal Model Context Protocol (MCP) Server for AI Agents
// Native integration for Antigravity, Cursor, Cline, Windsurf & Claude
{
"mcpServers": {
"sadasend": {
"url": "https://mcp.sadasend.com",
"headers": {
"Authorization": "Bearer $SADASEND_API_KEY"
}
}
}
}
// 12 email tools: send_email, preview_email, verify_email, get_stats, etc.Looking for full API schemas, error codes, webhook verifiers, or MCP tools?
Explore all documentationSeamless integration with the tools modern teams run on
Whether you dispatch from high-growth CRMs, visual automation pipelines, autonomous AI agent frameworks, or developer IDEs, SadaSend connects in seconds via REST API, Port 587 ESMTP, or native Model Context Protocol (MCP).
The leading modern open-source CRM (React/GraphQL). Route transactional updates, deal notifications, and customer emails with zero SMTP rate-limiting.
- Custom SMTP relay
- Zero Gmail rate limits
- Verified custom DKIM
The next-gen relational CRM for tech startups. Dispatch transactional customer sequences, onboarding alerts, and billing receipts via low-latency REST.
- 202 Accepted async queue
- Webhook event feedback
- Domain reputation isolation
Connect SadaSend as your dedicated transactional delivery engine for HubSpot workflows, password resets, and receipts with zero marketing stream bleed.
- Dedicated transactional stream
- STARTTLS encryption
- 100% SPF/DKIM alignment
Enterprise Service Cloud and CRM notification gateway. Enforce server-enforced recipient allowlists and tamper-proof delivery audit trails.
- Per-key recipient allowlists
- FCrDNS authenticated IPs
- Full delivery telemetry
Modern contact and relationship CRM. Send transactional messages and team notifications without risking your primary Google Workspace domain.
- Subdomain isolation
- Instant API key minting
- 6,000 free monthly sends
Automate transactional notifications, invoice dispatches, and alert workflows in n8n with zero SMTP socket timeouts or Gmail rate traps.
- Native HTTP request node
- Auto retry with DLQ
- Zero socket timeouts
Plug SadaSend into visual scenarios using the HTTP request module. High throughput, zero account freeze risks, and real-time webhook sync.
- JSON body templates
- Instant delivery tracking
- High burst capacity
Trigger transactional receipts, alerts, and invitations from 8,000+ apps. Deliver reliably with 202 Accepted durability and automated suppressions.
- Catch Hook triggers
- Structured metadata
- Bilateral webhook sync
Integrate in under 60 seconds across your stack
Standard SMTP Relay (Twenty CRM, HubSpot, Supabase)
host: smtp.sadasend.com
port: 587
security: STARTTLS
user: apikey
pass: $SADASEND_API_KEY # Scoped to recipient allowlistA hosted MCP server with built-in safety controls
Connect Claude, Cursor, Windsurf, or custom AI agents in one command. Your agent gets twelve safe tools to compose, preview, verify, and monitor email — with zero access to credential generation, domain deletion, or suppression tampering.
claude mcp add --transport http sadasend https://mcp.sadasend.com --header "Authorization: Bearer sada_agent_sk_…"| Tool | Scope | Does |
|---|---|---|
| send_email | send | Send one, honouring key mode and allowlist |
| preview_email | send | Dry run — render and validate, never send |
| get_email_status | read | Delivery state and full event trail |
| search_emails | read | Filtered log search — the debugging workhorse |
| list_domains | read | Which domains are verified — the first check when a send is refused |
| check_domain_setup | domains | Per-record DNS diagnosis, with the fix |
| list_templates | read | Templates on the account, with the variables each expects |
| check_suppression | read | Is this address suppressed, and why |
| render_template | read | Preview a template without sending |
| get_stats | read | Bounce and complaint rates for a window |
| build_utm_url | read | Build and normalize standardized UTM campaign URLs with validation |
| verify_email | read | Pre-flight MX DNS lookup, RFC syntax check, and disposable detection |
There is no create_api_key, no delete_domain, and no remove_suppression tool. An agent that can mint its own credentials or un-suppress a complainer has no safety limits. Those operations live exclusively in the dashboard under authenticated human control.
- “Send the welcome email to the three users who signed up this morning.”
- “Why did our password reset to dana@example.com bounce yesterday?”
- “Check whether our DNS is set up correctly for sending.”
- “Show me every email that bounced this week, grouped by reason.”
Ready to connect Antigravity, Cursor, Windsurf, or Claude Desktop?
Deliverability tools that tell you the fix
The same verification engine that onboards sending domains. Free, no sign-up, no email capture — up to 5 checks/day anonymously (or 100/mo free with an account).
- SPF checkerResolves every include and counts your lookups against the limit of ten.
- DKIM validatorFetches the selector, checks the key length, verifies a signature.
- DMARC checkerParses your policy and tells you what it actually does to failing mail.
- MX lookupResolves the chain and flags misconfigurations.
- Blocklist checkQueries the major blocklists for an IP or domain.
- Inbox placement testInteractive placement simulation across major mailbox providers.
- Email deliverability checkerReal-time syntax validation, MX lookup, typo autocorrect, and burner shield.
- Disposable email checkerNewDetects temporary, throwaway burner addresses before they bounce.
- All diagnostic toolsComplete directory of free deliverability utilities, syntax validators, and DNS checkers.
Where we excel, and where others focus
Every row in our column is a capability you can check against the live API today, not a roadmap. If you need a drag-and-drop newsletter builder for marketing campaigns, tools like SendGrid or Loops are built for that. But if you want to replace both legacy email APIs and costly list cleaners like NeverBounce with native pre-send shields, a built-in Verification Studio, server-enforced AI agent allowlists, and a hosted MCP server, choose SadaSend.
| Capability | SadaSend | Resend | SendGrid | Postmark | Mailgun | Loops |
|---|---|---|---|---|---|---|
| Transactional email API | Yes | Yes | Yes | Yes | Yes | Yes |
| Hosted MCP server | Yes | Yes | No | Partial | Partial | Yes |
| Per-key recipient allowlist | Yes | No | No | No | No | No |
| Agent approval mode | Yes | No | No | No | No | No |
| Dry run / simulation send | Yes | Partial | No | No | No | No |
| Scoped API keys | Yes | Yes | Yes | Partial | Partial | No |
| Per-key velocity & rate limit | Yes | No | No | No | No | No |
| Pre-send bounce shield | Yes | No | Partial | No | Partial | No |
| Built-in list verification studio | Yes | No | Partial | No | Partial | No |
| ESMTP relay (port 587 TLS) | Yes | Yes | Yes | Yes | Yes | Partial |
| Free deliverability tools | Yes | No | No | Partial | No | No |
| Zero-backend contact form API | Yes | No | No | No | No | No |
| Marketing campaign suite | No | Yes | Yes | No | Partial | Yes |
| Visual email designer | No | Yes | Yes | No | Yes | Yes |
Real engineering signals, zero synthetic testimonials
We removed fabricated logos and paid endorsements. Instead, inspect live API latency measured directly from your browser, install our open SDKs from standard registries, and audit transparent deployment history.
Round-trip time from your browser to the SadaSend API in Falkenstein, Germany, measured when this page loaded.
The status page checks the REST API, hosted MCP, dashboard and site when you open it. No uptime percentage is published until a monitor has recorded one.
Every release passes typecheck, the unit and end-to-end test suites, and a database snapshot before it is deployed.
Enterprise-grade safety on an indie budget
Every guardrail and all twelve MCP tools are included on every tier, including the free plan. We do not gate safety behind expensive enterprise sales calls or force multi-tenant users onto degraded, unmonitored pools.
- Full Model Context Protocol (MCP) server (Claude, Cursor & AI runtimes)
- Every agent guardrail: Scoped keys, recipient allowlists, approval mode & dry run
- 6,000 emails a month, up to 200 a day
- 2 active Contact Form endpoints with invisible honeypot spam shield & submissions inbox
- 100 pre-send email verifications / mo (real-time syntax, MX & burner checks)
- 3 verified sending domains with automatic SPF, DKIM & DMARC diagnosis
- Official zero-dependency SDKs (TypeScript & Python) + REST API & SMTP
- 30-day message & log retention
- Single-seat developer workspace (expandable with paid plans)
- Everything in Free, with complete agent safety guardrails
- 10 active Contact Forms with automated customer confirmation auto-responders & CSV lead exports
- Slack 1-Click Interactive Approvals (review & release held agent emails in Slack)
- 15,000 emails a month, with no daily cap once warmup completes
- 1,000 pre-send email verifications / mo + Batch CSV list cleaner (100 / batch)
- 6 verified sending domains with automated DNS diagnostics
- 3 signed webhook endpoints with signature verification & event replay
- 25 versioned Liquid templates
- 30-day message & log retention
- Optional overage at $0.60 per 1,000 (off until you raise Spend Guard) or never-expiring credit packs
- Free white-glove migration assistance from legacy providers
- Everything in Starter
- 50 active Contact Forms with custom redirect URLs & CSV export
- 75,000 emails a month
- Inbound MX Routing & Token-Optimized Webhook Parsing (~85% LLM token savings)
- Native DMARC Aggregate XML Forensic Radar (RFC 7489 ingestion)
- 10,000 pre-send email verifications / mo + Batch CSV list cleaner (1,000 / batch)
- 25 verified sending domains
- 60-day message & log retention
- 10 signed webhook endpoints with windowed dead-letter replay
- 100 versioned Liquid templates
- Optional overage at $0.50 per 1,000 (off until you raise Spend Guard) & prepaid credit packs
- Developer support via in-app tickets, answered by an engineer
- Everything in Pro
- Unlimited active Contact Forms with custom subject templates & multiple recipients
- 250,000 emails a month
- Inbound MX Routing & Token-Optimized Webhooks (all active domains)
- Native DMARC Aggregate XML Forensic Radar & Automated Policy Scorecard
- 50,000 pre-send email verifications / mo + High-speed batch cleaner (10,000 / batch)
- Unlimited verified sending domains
- 90-day message & log retention
- 25 signed webhook endpoints with dead-letter replay
- Unlimited versioned templates
- Optional overage at $0.40 per 1,000 (off until you raise Spend Guard) & high-volume credit packs
- Developer support via in-app tickets, answered by an engineer
Start on the free plan without a credit card. Yearly billing saves 15% on Starter and Pro, 20% on Scale. Transparent overage: $0.60, $0.50, or $0.40 per thousand emails by plan, billed only for what you send. Zero per-seat charges, ever.
Straight answers
Each answer stands on its own, so you can quote it without the rest of the page.
https://mcp.sadasend.com and your agent instantly receives twelve safe tools: send, preview dry run, email search, delivery status check, template rendering, domain verification, and DNS diagnostics. Destructive operations like creating API keys, deleting domains, or clearing suppressions are architecturally absent from the MCP server, ensuring your agent cannot exceed its mandate.*@yourcompany.com). Any send attempt outside the allowlist is refused synchronously at the API boundary before entering the delivery queue. For sensitive workflows, enable human approval mode to hold emails in a pending_approval state until verified via your dashboard or webhook.smtp.sadasend.com on port 587 with STARTTLS, using your API key as the password. Every guardrail—scopes, allowlists, suppressions, and rate limits—is enforced on SMTP identically to our REST API. Migrate in seconds without changing your application code.npm install sadasend / bun add sadasend) and Python (pip install sadasend) SDKs are live on npm and PyPI with zero third-party runtime dependencies. They include typed exceptions, automatic idempotency keys, and server-enforced agent guardrails. You can also integrate via standard REST API (fetch/cURL), SMTP on port 587, or our hosted Model Context Protocol (MCP) server for AI agents.Stop worrying about runaway email. Ship today.
Generate a scoped API key, set your recipient allowlist, and point your agent or backend at SadaSend. 6,000 emails a month free, with no credit card required.
- 6,000 emails a month, free forever
- All 12 MCP tools included
- Every guardrail active on Day 1
- No credit card required