One flat-fee platform on your own SES vs two metered products

SendGrid is huge and battle-tested — it claims to deliver over 200 billion emails a month. This comparison isn't about whether SendGrid works. It's about whether its model still fits you. It's most useful for three kinds of teams:

• teams pushed off SendGrid's discontinued free tier and looking for a durable free home
• teams paying for both Email API and Marketing Campaigns and questioning the two meters
• teams that want email infrastructure they own — in their AWS account, on their reputation

Domain authentication

The missing layer: who else is sending as your domain?

A SendGrid dashboard answers one question well: what happened to the mail you sent through SendGrid. It cannot see the billing system, the helpdesk, or the analytics tool that still puts your domain in the visible From address. DMARC aggregate reports can, because mailbox providers report on every service that authenticates — or fails to authenticate — as that domain.

Domain-based Message Authentication, Reporting, and Conformance (DMARC) covers the exact domain in the visible From address and every service sending as it. It does not cover lookalike domains that someone else registers.

SendGrid documents DMARC setup and points to Valimail for monitoring and enforcement. Valimail is a mature dedicated DMARC product, and for some teams a specialist vendor is exactly the right answer. It is also a hand-off: the platform sending your mail and the tool reading your reports are two different vendors.

SendOps takes the other approach and puts DMARC monitoring inside the platform that operates your SES. Each verified domain gets its own aggregate reporting address, and the senders that appear are grouped by provider fingerprint, network, and behaviour. Because SendOps knows which DKIM selectors are yours, it can prove your own SES sending and separate a first-party misconfiguration from traffic you have not yet recognised.

An unrecognised source is one that needs a decision, not an attacker. It might be a forgotten SaaS product, a shared sending pool, or ordinary forwarding.

Rejecting a source records that it is not yours. It does not block the source or silence future alerts. A stricter DMARC policy is what asks receiving providers to act on unauthenticated mail.

When the evidence supports it, SendOps shows what p=quarantine or p=reject would affect in messages per day, alongside the exact record and a field-level diff. You publish that record in your own DNS provider. SendOps never writes DNS and never advances your policy for you.

Pricing model

Two products, two meters

SendGrid splits email into two products. The Email API handles transactional sending, with list prices from $19.95/mo (Essentials) to $89.95/mo (Pro, which includes a dedicated IP), metered by send volume. Marketing Campaigns is a separate product with its own meter — Basic from $15/mo for 5,000 contacts and 15,000 sends, Advanced from $60/mo, scaling with contact count and capped sends.

That means a team doing both transactional and marketing email carries two plans, two meters, and two upgrade paths — one driven by send volume, the other by contact count. Growth on either axis moves a bill.

SendOps is one platform with one flat fee. Contacts, SendQL segments, broadcasts, SendFlow drip workflows, templates, and deliverability analytics are all included, on every plan. Delivery isn't metered by SendOps at all — AWS bills you directly at SES rates, $0.10 per 1,000 emails.

The free tier

SendGrid removed its free plan. Ours has no metered volume.

For years, SendGrid's free tier was the default starting point for developer email. It has since been discontinued — new users start at paid plans, and many teams searching for a SendGrid alternative cite exactly this change.

SendOps Free costs $0 and has no metered volume and no contact limits. That's possible because of the architecture: SendOps doesn't resell delivery, so there's no per-email cost for us to recover. You pay AWS directly for what you send — about $0.10 per 1,000 emails through SES — and the platform layer is free until you need team features.

A free plan built on your own infrastructure also can't be taken away the same way. If SendOps changed its pricing tomorrow, your SES account, your domains, and your reputation would still be yours.

Cost at volume

The cost curve at a million emails a month

At around one million emails a month, SendGrid's Email API list pricing typically runs several hundred dollars — before adding Marketing Campaigns, which is metered separately by contacts and sends. Exact numbers depend on plan, overages, and negotiated rates, so treat published tiers as a floor.

The same million emails through SES costs about $100 in delivery. SendOps adds a flat $29/mo (Team) or $149/mo (Business) on top — and that fee doesn't move with volume or contact count, because SendOps never meters your sending.

The structural point matters more than any single price snapshot: with SendGrid, cost scales with usage on two meters. With SendOps, delivery scales at SES rates and the platform fee stays flat.

Event history

Seven days of history isn't an audit trail

On some SendGrid plans, email activity history is retained for as little as 7 days. If a customer asks "did you email me last month?", or compliance needs a record of a notification from a quarter ago, a one-week window means building your own event archive on top of webhooks.

SendOps retains email activity for up to a year, searchable in the dashboard and queryable through the activity API. Deliverability analytics — bounces, complaints, engagement by domain — are built on the same history, so trends are visible over months, not days.

And because the underlying events come from SES in your own account, the raw data is yours to keep for as long as you want, independent of any SendOps retention window.

Ownership and reputation

Your account, your reputation, your data

On SendGrid's shared IP pools, your deliverability is partly shaped by other senders' behavior — a dedicated IP that isolates your reputation is included from the Pro plan at $89.95/mo list. With SendOps, isolation is the default, not an upgrade:

This also changes what leaving looks like. Migrating off SendGrid means moving sending infrastructure, contact data, and reputation. Leaving SendOps means turning off a control plane — your SES account, domains, and history stay exactly where they were.

Everything as code

Built for git workflows — and for agents

SendGrid's marketing automation lives in its dashboard: templates in a web editor, journeys built by clicking. That works for a human with a mouse, but it isn't reviewable, versionable, or operable by software — an agent can't click through a journey builder, and WYSIWYG output is an opaque blob it can't read or edit.

SendOps treats email as code. Template source is React Email TSX in your repository, compiled by SendOps to SES-native Handlebars on deploy — and edits made in the visual editor come back as pull requests. Segments are SendQL queries. Drip workflows are SendFlow definitions with both a DSL and a visual canvas, so engineers and marketers edit the same artifact. Assets ship through a CDN, and tracking domains are configured infrastructure, not dashboard settings.

Because every definition has a textual representation, coding agents can draft a template, open the pull request, define a segment, and wire a workflow the same way your team does. A human approves the merge; agents cannot.

Definitions are not customer data. Contacts, list membership, consent, activity, and delivery results stay in SendOps, where an authorized agent reaches them through the native MCP server or the Public API, which clients authorize with OAuth 2.1 rather than copied long-lived keys.

The guarded operations stay guarded. Sending a broadcast and activating a workflow each require an explicit confirmation, and a workflow run can be previewed as a rendered email before it sends.