SendPost Review 2026: A Free Plan on One Page, a Demo Wall on the Other

The Best: A working SendGrid-compatible endpoint and four documented SMTP ports

The Worst: No price published anywhere, and two vendor pages contradict each other on the free plan

Compare SendPost with 290+ platforms Read pricing Starting from On demand (upon demo)
G2G2 4.7/5 (3 reviews, none rendered)
|
CCapterra 5.0/5 (3 reviews, all read) Six reviews exist in total
|
Trustpilot N/A (no listing)
At a glance Vendor documentation read twice + the complete 3-row Capterra population + 43-repository GitHub enumeration, August 2026
LevelIntermediate (developer-first)
Best forTeams migrating off SendGrid who want to change a URL, not their code
Strongest pointA SendGrid-compatible endpoint and four fully documented SMTP ports
Weakest pointNo price published anywhere, and two vendor pages contradict each other
Ban toleranceNot rated, no published industry list found (terms of Oct 2020 ban purchased lists only)
Refund policyOnly if SendPost terminates without cause mid-term, none after an abuse termination
TransactionalYes (primary)
SMTP RelayYes (four ports)
APIYes (SendGrid-compatible)
PromotionalNo
AutomationNo
MCPNo

The short version

SendPost is the smallest platform on this site by public evidence. Three Capterra reviews, three G2 reviews with none of them rendered, no Trustpilot listing, no Product Hunt page, no Reddit thread we could find. That is six reviews in total, of which we were able to read three.

What is there is technically solid. The SMTP reference publishes a hostname, four ports and the encryption mode for each, which is more than Campaign Monitor, Selzy or Sendloop manage. The API publishes a base URL, an authentication header and real endpoints. There are 43 public repositories on GitHub, including fourteen framework example projects all updated on the same January 2026 day. There is a SendGrid-compatible endpoint that lets you point an existing integration at SendPost by changing a URL.

What is not there is the thing a buyer needs first. SendPost does not publish a price. Not a plan name, not a tier, not a number. The pricing page shows a Schedule Demo button in three places and nothing else. And on the two vendor pages that answer the question “Do you offer a free plan or trial?”, the answers are opposite.

Our take

A credible engineering product with almost no public record and a commercial front door that contradicts itself. If you are migrating off SendGrid and you like the team, book the demo and ask for the numbers in writing. If you want to compare prices before you talk to anyone, this is not a platform you can evaluate.

Two answers to the same question

SendPost is marketed from two surfaces: its own site at sendpost.io, and a set of pages hosted on sendx.io. Both carry a FAQ. Both FAQs ask the same question. The answers do not match.

PageQuestionAnswer, quoted literally
sendx.io/sendpostDo you offer a free plan or trial?“No, at the moment we require a demo and up front financial commitment to use SendPost.”
sendx.io/sendpost/features/email-apiDo you offer a free plan or trial?“Yes, we offer a very generous free forever plan that includes up to 25,000 free emails every month.” Followed by: “No credit card is required to sign up.”

We read both pages twice, on the same URLs, demanding the literal surrounding text on the second pass. Both strings are real. On the first page, the words free forever and 25,000 do not appear at all.

Which one is current?

The canonical pricing page, www.sendpost.io/pricing, contains no plan name, no quota and no dollar figure of any kind. It offers a demo. On the evidence of the page the vendor itself designates as its pricing page, the demo-and-commitment answer is the operative one, and the free forever plan is a claim left standing on a feature page. We are not publishing 25,000 free emails a month as a fact about SendPost.

Capterra’s listing shows “Starting price $7 Usage Based, Per Month”. That figure appears nowhere in SendPost’s own materials. It is third-party reported and we do not repeat it as a price.

Snapshot

Legal entitySendPost, LLC, “doing business as SendWorks and SendX”. The marketing says SendPost is a product of SendWorks; the terms of service say the company is called SendPost and SendWorks is one of its trading names. The hierarchy is inverted.
Address1580 N. Logan Street, Suite 600, PMB 38591, Denver, Colorado. PMB means private mailbox. No operational address is published.
Governing lawDelaware. Terms last updated 31 July 2026.
Product family“products currently or previously offered under the names SendWorks, SendX, SendPost, SendVerify, and SendHealth”
Origin“SendPost was built by the team of engineers and deliverability experts at SendX, which processes over 300 million emails/month.”
Founding yearNot published for SendPost. SendX states it was founded in 2017; that is a different product.
FoundersNot published for SendPost. SendX names Trevor Hathfield as CEO, Varun Jain as CTO and Joshua Waldman as COO, none of them labelled founder.
PriceNot published. The pricing page carries no figure.
ComplianceNo certification published. SOC 2, ISO 27001, GDPR, CCPA, HIPAA and PCI are all absent from the features page and the security page.
Status pageNone. status.sendpost.io does not resolve.

The SMTP setup, which is the best documented thing here

This is where SendPost is unambiguously good, and it is worth saying plainly because so few providers at this size are. The API reference publishes a hostname, every port, and what each port is for.

SettingValue
Hostnamesmtp.sendpost.io
Unencrypted or TLSPorts 25, 2525 or 587, quoted literally: “For an unencrypted or a TLS connection, use port 25, 2525 or 587.”
SSLPort 465: “For a SSL connection, use port 465”
STARTTLS“SendPost supports STARTTLS for establishing a TLS-encrypted connection.”
Recommended587
AuthenticationUsername and password taken from the account. The AUTH mechanism is not named.
Messages per connectionNot published.
Rate limitNot published. The features page mentions “Rate limiting & abuse prevention mechanisms” with no figure attached.

We read that page three times because we have been burned once already in this project by a summariser inventing a table of SMTP settings that did not exist. These values reproduce literally on every read.

The API, and the two things it does not publish

The base URL is https://api.sendpost.io/api/v1, corroborated independently by the README of the JavaScript SDK. Sub-account authentication uses the header X-SubAccount-ApiKey. Sending is a POST to /subaccount/email/. Real-time webhooks are documented.

Two things are missing and both matter.

There are no published rate limits. Not for the API, not for SMTP, not per sub-account. For a transactional provider this is the number an engineer needs before writing a retry policy. MailerSend publishes every one of its limits; SendPost publishes none.

There is no MCP server. We enumerated all 43 public repositories and none carries the name. Nothing in the documentation references the protocol.

The migration APIs, one confirmed and one that is not what it sounds like

The SendGrid-compatible API is real. It lives at https://api.sendpost.io/v1/sendgrid/v3/mail/send, described as “Migrate from SendGrid with minimal code changes using SendPost’s SendGrid-Compatible API.” If you have a SendGrid integration in production, that is a genuine one-line move, and given what has happened to SendGrid accounts without notice it is a sensible thing to build.

An Amazon SES compatible API is often attributed to SendPost and we could not confirm it. The documentation path for it returns 404, and the string does not appear on the SES page. What exists is an integration that routes through an IP pool parameter. That is not the same thing, and we are not describing it as one.

SDK count, which the vendor gives two answers to

The documentation says “SDKs in 6 programming languages”, and six is what the current repositories show: JavaScript, Python, Go, Ruby, Java and PHP. The pricing page says “SDKs in 14+ programming languages”. Both are on vendor surfaces. Six is the number we can count.

A detail worth noting for anyone integrating: the documentation writes the account-level header as both X-Account-APIKey and X-Account-ApiKey on the same page, and the example API key is printed in the clear. Neither is dangerous. Both are the kind of thing a documentation pass catches.

Is anyone still building it

Yes, and the GitHub record is the clearest evidence in this review because it is dated and public.

DateWhat happened
January 2025 to March 2025The old generation of SDKs is archived one language at a time: ruby, javascript, rust and php in January, python in January, go in March. The Rust SDK is archived and not replaced, so Rust is a dropped language.
26 December 2025A new generation appears: sendpost-ruby-sdk, sendpost-python-sdk, sendpost-go-sdk, sendpost-java-sdk, sendpost-js-sdk, plus ESP example projects.
29 December 2025Seven more example repositories and the PHP SDK.
15 January 2026Fourteen framework example projects all updated on the same day: Spring Boot, Rails, Laravel, Supabase Edge Functions, Netlify, Vercel, Cloudflare Workers, NestJS, Django, AWS Lambda, Next.js, FastAPI, Flask and Express.
16 January 2026 to todayNothing visible. Roughly seven and a half months with no public commit activity.

That is a real, funded engineering push followed by a long quiet stretch. It is not the profile of an abandoned product, and it is not the profile of one under continuous development either. The old java SDK, incidentally, was never archived alongside its five peers, which is the sort of loose end that tells you the clean-up was done by hand and not finished.

Three reviews, and what we can honestly say from them

Capterra shows 5.0 from 3 reviews. We read all three. G2 shows 4.7 from 3 reviews and renders none of them, so we read zero of those and can quote nothing from G2. Trustpilot has no listing for sendpost.io. Product Hunt has no page.

What a three-review corpus can and cannot tell you

Nothing statistical. A 5.0 average across three reviews is not a quality signal, it is an absence of dissatisfied users who wrote something down. We report the three because they are all the primary testimony that exists, and because two of them name something specific.

DateReviewerStarsWhat they said
14 May 2024Paul W., Consultant, Textiles5.0Pros: “Easy peasy onboarding. Super speedy UX. Great support!” Cons: “No complaints. Never looked back. So far, so good.” Switched from ActiveCampaign.
19 April 2024Giorgi S., COO, Gambling and Casinos5.0Pros: “They are great in support and resolving the issues super fast”. Cons: “System is tracking duplicated open rates”. This is the only substantive complaint in the entire corpus.
28 July 2023Aaron K., SaaS Onboarding Specialist, IT and Services5.0Titled “The SendGrid alternative we dreamed of”. The pros and cons fields are truncated in the listing and did not reproduce identically across two reads, so we do not quote them.

Support is named positively in two of the three rendered rows. Company size is displayed for none of them. The newest review is from May 2024, which makes the corpus roughly twenty-seven months old.

One number we checked and are not publishing

Capterra appeared on a first read to show a Features sub-score of 5.0. It did not reproduce on the second read of the same page. Under our own rules that makes it unverified, so it is not in the table below. Ease of use, value for money and customer service all reproduce at 5.0 from 3.

Deliverability, where every number is the vendor’s own

SendPost is not in the EmailToolTester seedlist panel, which was frozen in January 2024 and has since been discontinued in favour of feature assessments. It has no page on emaildeliverabilityreport.com. There is no independent measurement of SendPost inbox placement anywhere that we could find.

What the vendor publishes about itself:

  • “SendPost maintains 95-98% delivery rates and maintained its deliverability team size at only 3 people.”
  • “95-98% inbox placement even as volume scaled from 100M to 300M emails per month”

Those are two different metrics. Delivery rate counts messages the receiving server accepted. Inbox placement counts messages that landed in the inbox rather than the spam folder. They are never the same figure, and a provider that quotes one identical range for both has not measured them separately.

Volume is quoted three different ways across the same family of pages: “Sending over 413+ million emails per month”, “over 300 million emails/month” for SendX, and “100M to 300M” in the case study. These may all be true of different entities at different dates. Read together on one afternoon, they are three numbers for one claim.

The infrastructure story is better than the numbers. Shared or dedicated IPs, automatic IP warmup, real-time bounce and complaint handling, alerting, and a deliverability team that came out of an ESP doing serious volume. That is why the score below is 5.5 and not lower. It is not 7 because nothing here has been checked by anyone outside the company.

What is not published at all

Collected in one place, because for a provider aimed at businesses these absences are the buying risk:

ItemStatus
Price, any planNot published
API and SMTP rate limitsNot published
Compliance certification (SOC 2, ISO 27001, HIPAA, PCI)None published
GDPR or CCPA statement on the security pageAbsent
EU or US region choice, data residencyNot published
Suppression list handlingNot published
Status pageDoes not exist
Changelog or release notesDoes not exist
Founding year and founders of SendPostNot published

Log retention is published, and it is the one operational number they do give: “Email Event Logs Uncover exactly what happened to your emails – with full transparency, search, and retention for up to a year.” Note the words up to. That is a ceiling, not a guarantee, and it is almost certainly plan-dependent, which we cannot check because there are no plans published.

Ratings breakdown

Ease of use
6.5
Automation
5.5
Deliverability
5.5
Support
6.5
Pricing value
4.5
API and MCP
6.5
Overall5.8 / 10
How we score

Six categories, each out of 10, judged against the same evidence rules on every platform we cover. The overall is the unweighted arithmetic mean, rounded to one decimal: 6.5 + 5.5 + 5.5 + 6.5 + 4.5 + 6.5 = 35.0, divided by 6, equals 5.8.

Ease of use, 6.5. The documentation is clear, there are fourteen framework starters to copy from, and the two usable reviews both praise onboarding. Held down because you cannot sign up and try it: the front door is a demo request.

Automation, 5.5. This is a transactional relay, not a marketing platform, and it is scored as one. Templates with Handlebars, AMP support, groups, sub-accounts and real-time webhooks are all present. There is no campaign builder, no segmentation, no journey editor, and none is claimed.

Deliverability, 5.5. Good infrastructure, an experienced team, dedicated IPs with automatic warmup. Zero independent verification, two incompatible metrics quoted with one range, three different volume figures, and no compliance certification of any kind.

Support, 6.5. Named positively in two of the three rendered Capterra rows, one of them specifically for speed of resolution. Held down because the corpus is three reviews old enough to be historical, there is no status page, and no support SLA is published.

Pricing value, 4.5. The lowest score we have given in this category. There is no price. There is no plan. The one page that promises a free forever tier is contradicted by the page the vendor calls its pricing page. You cannot assess value against an unpublished number, and a buyer should not have to book a call to learn the shape of the offer.

API and MCP, 6.5. Base URL, auth header, webhooks, six SDKs, a working SendGrid-compatible migration endpoint and a genuinely complete SMTP reference. Held down by unpublished rate limits, no MCP server, an SES-compatible API that does not exist as described, and inconsistent header casing in the docs.

Who should use SendPost, and who should not

SendPost is the right call for

Teams migrating off SendGrid in a hurry
The SendGrid-compatible endpoint is the single best reason to look here. Point your existing client at a different URL and keep the code. One reviewer titled their review exactly that.
Platforms sending on behalf of their own customers
Sub-accounts are built into the authentication model rather than bolted on, with a dedicated multi-tenant guide and per-sub-account API keys.
Senders who want a real conversation before committing
The demo wall is a genuine advantage if you were going to negotiate anyway. Small vendor, direct access, and pricing that is evidently bespoke.

Look elsewhere if

You need to compare prices before you talk to a salesperson
There is nothing to compare. Postmark, MailerSend and SMTP2GO all publish a full price list you can put in a spreadsheet this afternoon.
Procurement will ask for a SOC 2 report
No certification of any kind is published, and neither is a data residency option. Ask before you scope the project, not after.
You need published rate limits to design around
None exist. MailerSend publishes every limit it enforces, down to the API window, which is why it scores higher here on the same category.
You want a track record you can read
Six reviews in total, the newest from May 2024, no status page, no changelog, and no public activity since January 2026. That is not evidence of a problem; it is an absence of evidence either way.

Third-party validation

SourceScoreVolumeNotes
Capterra5.0 / 53 reviewsEase of use 5.0, value for money 5.0, customer service 5.0, each from 3. Product ID 218757, verified to resolve to SendPost. All three rows read.
G24.7 / 53 reviewsAggregate shown as of 17 August 2026. No individual review is rendered, so we read none and quote none.
TrustpilotN/ANo listing for sendpost.io. A listing exists for sendx.io and is a different product.
Product HuntN/ANo page.
Inbox placementNo figureAbsent from the EmailToolTester panel and from emaildeliverabilityreport.com.

We never add Capterra and G2 into a single denominator. Where this review says something is named by reviewers, it means named in that many of the three rendered Capterra rows.

Top 3 alternatives

AlternativeOur scoreTake it instead if
Postmark7.1You want the same transactional job done by a provider with a published price list, a status page, a public incident history and a corpus large enough to mean something.
MailerSend7.1You want every rate limit and quota printed on a page before you write a line of code. It is the direct answer to SendPost’s biggest documentation gap.
SMTP2GO7.5You want the highest-rated relay we cover, with a hostname and port set published per region and the best review corpus in the category.

Continue reading

Frequently asked questions

What is SendPost?

A transactional email API and SMTP relay, marketed as a product of SendWorks and built by the engineering team behind SendX. The legal entity is SendPost, LLC of Denver, Colorado, doing business as SendWorks and SendX, governed by Delaware law. It is aimed at developers and at platforms that send on behalf of their own customers, not at marketers building campaigns.

How much does SendPost cost?

We cannot tell you, because SendPost does not publish a price. Its pricing page contains no plan name, no quota and no dollar figure, only a request to schedule a demo. One vendor FAQ states that a demo and an up front financial commitment are required. Capterra shows a starting price of $7 per month on a usage basis; that figure appears nowhere in SendPost’s own materials and we do not repeat it as fact.

Does SendPost have a free plan?

The vendor gives two opposite answers to that exact question on two of its own pages. One says no free plan or trial is offered and a demo with an up front commitment is required. The other says there is a free forever plan of up to 25,000 emails a month with no credit card. The canonical pricing page publishes neither. Treat the free tier as unconfirmed and ask for it in writing.

What are SendPost’s SMTP settings?

The hostname is smtp.sendpost.io. Ports 25, 2525 and 587 are for unencrypted or TLS connections, with STARTTLS supported, and port 465 is for SSL. The vendor recommends 587. Authentication uses the username and password from your account; the AUTH mechanism is not named, and no per-connection message limit or rate limit is published anywhere.

Is SendPost a good SendGrid alternative?

Technically it is set up to be one. There is a SendGrid-compatible endpoint at api.sendpost.io/v1/sendgrid/v3/mail/send, described by the vendor as letting you migrate with minimal code changes, and one of the three Capterra reviews is titled exactly that. The reservation is commercial rather than technical: SendGrid publishes its prices and SendPost does not.

Is SendPost still being developed?

Yes, though not continuously. Its GitHub organisation holds 43 public repositories, with an entire SDK generation rebuilt in late December 2025 and fourteen framework example projects all updated on 15 January 2026. Nothing public has changed since then, roughly seven and a half months. There is no status page and no changelog, so GitHub is the only dated liveness signal available.

Changelog

DateChange
27 August 2026First publication. All 3 rendered Capterra reviews read, the complete population; G2 renders none of its 3 and is reported as unreadable rather than summarised. Every SMTP port, hostname, header and quoted string verified on a second literal read of the same URL. A Capterra Features sub-score that did not reproduce on the second read was discarded. The two contradictory free-plan answers were each confirmed twice on their own URLs. An Amazon SES compatible API widely attributed to SendPost was searched for and not found, and is reported as unconfirmed rather than repeated. The $7 starting price shown by Capterra is not published by the vendor and is not repeated here as a price.
A

Alaa Touil

Founder, SMTPedia

Runs the SMTPedia editorial testing lab. 10+ years hands-on with SMTP infrastructure, deliverability protocols, and ESP evaluation. Every review is based on his direct testing on real sending accounts.

Read more →
R

Rabeb

Reviewer, SMTPedia

Reviews all SMTPedia platform verdicts for editorial accuracy, sourcing rigor, and fair representation across price tiers and use cases.

Read more →

This review follows our email infrastructure testing methodology. We disclose affiliate relationships in our editorial independence policy.