Email marketing MCP free tier
If you want to test subject lines from inside an AI agent without paying upfront, marginal's free tier gives you 100 experiments every month on a fully hosted email MCP server. Here's what's included and how to get started.
What the free tier covers
The email marketing MCP free tier on marginal includes 100 experiments per month. An experiment is one subject line A/B test: you generate variants, launch a managed send, and read back open and click metrics with within-test lift figures.
- 100 experiments/month at no cost
- Subject line A/B tests with open and click tracking
- Within-test lift metrics to compare variants
- Managed sends — marginal handles delivery
- Bearer API key auth (marg_live_...) for your agent
Why use a hosted email MCP server
marginal is a hosted email MCP server, so there is nothing to deploy or maintain. You point your AI agent at https://marginal.sh/mcp, add your API key, and the four tools become available immediately.
Because it runs as a managed service, the free tier behaves identically to paid usage — same endpoint, same tools — just capped at 100 experiments per month. That makes it a low-risk way to evaluate AI agent email marketing before committing.
- generate_variants — draft subject line candidates
- launch_test — start a managed A/B send
- get_results — pull opens, clicks, and lift
- recommend_next — decide the follow-up step
Which agents can connect
Any MCP-capable client can use the free tier. That includes Cursor, OpenAI Codex, Claude Desktop, Claude Code, Windsurf, Cline, Continue, and Zed. You drive subject line A/B tests through natural language while the agent calls the underlying tools.
- Cursor, Claude Desktop, Claude Code
- OpenAI Codex, Windsurf, Zed
- Cline and Continue
- Registry id: sh.marginal/mcp
Getting the most from 100 experiments
On a free tier, each experiment counts, so be deliberate about what you test. Focus a single A/B test on one clear hypothesis — a tone shift, a personalization token, or a length change — rather than testing everything at once.
Use the results loop to compound learnings: read get_results, let recommend_next suggest the strongest direction, and roll that into your next variant batch. See the docs at https://marginal.sh/docs/ for the full tool reference.
- Test one variable per subject line A/B test
- Let recommend_next guide your follow-up
- Track lift, not just raw open rate
- Reserve experiments for sends that matter
Get started with marginal
marginal is a hosted email marketing MCP server at marginal.sh. Sign up free, create an API key, and connect https://marginal.sh/mcp from Cursor, Codex, or Claude Desktop.
- 100 experiments/month on the free tier
- Four MCP tools: generate_variants, launch_test, get_results, recommend_next
- Listed in the MCP Registry as sh.marginal/mcp