Email API Pricing Models: Predict Your Costs

Command references used in this guide: nylas demo email list, nylas init, and nylas auth list.

What pricing models do email APIs use?

Email APIs use 3 pricing models: per-message, per-connected-account, and usage-based. Per-message pricing charges for each email sent from your domain. Per-connected-account pricing charges a monthly rate for each mailbox the API can read and write. Usage-based pricing meters consumption of a specific resource, like hours of recorded meetings. Which model a vendor uses tells you what the product actually does.

The models aren't interchangeable because the products aren't. A transactional API like Amazon SES sends from your domain and never sees a user's inbox, so metering messages makes sense. A contextual API like Nylas connects to mailboxes users already have (Gmail, Outlook, Exchange, and others) for two-way sync, so the natural unit is the account. The email API comparison guide covers that product split in depth; this page covers what each model does to your bill.

How does per-connected-account pricing work?

Per-connected-account pricing charges a base monthly fee plus a rate for each mailbox beyond an included allowance. On the Nylas pricing page (June 2026), the Full Platform plan is $15/month including 5 connected accounts, then $2 per additional account per month. A calendar-only plan runs $10/month including 5 accounts, then $1.50 per additional account.

The math stays linear as you grow, which makes forecasting a spreadsheet one-liner. At 5 accounts you pay the $15 base. At 100 accounts: $15 + (95 × $2) = $205/month. At 1,000 accounts: $15 + (995 × $2) = $2,005/month, before the volume discounts that enterprise plans negotiate. The number that doesn't appear anywhere in that math is message volume: a connected account that syncs 10,000 emails costs the same as one that syncs 50, so a usage spike never spikes the bill.

How does per-message pricing work?

Per-message pricing meters each email sent, either as a flat rate per thousand or as monthly tiers with included volume. Public pricing as of May 2026: Amazon SES charges $0.10 per 1,000 emails with no monthly minimum; SendGrid starts at $19.95/month for 50,000 emails; Postmark starts at $15/month for 10,000.

This model rewards predictable broadcast volume and punishes spiky or conversational workloads. Sending 50,000 receipts a month costs $5 on SES; that's hard to beat. But the model only covers outbound mail from your own domain. Reading a customer's inbox, syncing their calendar, or sending as them isn't a feature you can buy at any per-message price from a transactional provider, which is why the two models coexist instead of competing.

Which pricing model fits which use case?

Match the model to the direction your email flows. Outbound-only from your domain points to per-message pricing; two-way mailbox access points to per-account pricing; metered features like meeting recording point to usage-based. The table below pairs common workloads with the model that bills them sanely.

Some products genuinely need both: a SaaS app might send receipts through SES at $0.10 per 1,000 while syncing support inboxes through a per-account API. Running two vendors for two jobs is normal, and usually cheaper than forcing one model to cover both.

How do you estimate costs before committing?

Per-account pricing makes estimation concrete: count the mailboxes, apply the rate card, done. The free sandbox tier supports up to 5 connected accounts for building and testing, so a working prototype costs $0. The nylas demo email list sample command goes one step earlier, exploring the CLI against realistic sample data with no credentials at all.

# Explore with sample data — no account, no credentials
nylas demo email list

# Set up the real thing (free sandbox, up to 5 accounts)
nylas init

# Count the accounts you've actually connected
nylas auth list

The 5-account allowance covers a realistic pilot: your own mailbox plus a handful of teammates exercising real Gmail and Outlook data for a few weeks. When the pilot graduates, nylas auth list is your billing forecast — multiply the accounts beyond 5 by $2 and add the $15 base.

What hidden costs should you check?

The rate card is rarely the whole bill. Four line items move real money and only appear in the fine print: SLA eligibility, support tiers, overage behavior, and contract-length requirements. The Nylas pricing page is candid about the first two, which is the pattern to look for in any vendor: the 99.99% availability SLA applies to annual contracts, and phone support with a dedicated CSM is a paid add-on rather than a default.

1. SLA eligibility — guarantees often require an annual contract. Month-to-month plans may carry no SLA at all (see the SLA guide for what the tiers mean).
2. Support tiers — email-during-business-hours is the usual baseline; severity-based response times and phone support are add-ons.
3. Overage behavior — does account #6 get billed automatically, or does the API start rejecting? Automatic billing is friendlier to production and worse for surprise invoices.
4. Deliverability extras — per-message vendors charge for dedicated IPs and domain warm-up tooling that the headline tier excludes.

One number worth writing into any vendor comparison spreadsheet: the cost at 10× your launch scale. Linear models like per-account pricing make that cell trivial to fill; tiered per-message plans sometimes jump sharply at the tier boundary, and that jump is better discovered in a spreadsheet than on an invoice.

Next steps

• Best email API for developers — the full 6-vendor comparison this pricing breakdown plugs into
• Email API SLAs: uptime vs success rate — what the annual-contract SLA actually guarantees
• Getting started — install the CLI and connect your first sandbox account
• SendGrid vs Nylas — a head-to-head of the two pricing models in practice
• Full command reference — every flag and subcommand documented