Frequently Asked Questions

Forward Email is a fully featured email service provider and email hosting provider for custom domain names.

We're an all-in-one alternative to Gmail + Mailchimp + Sendgrid.

Think of us as the service that can power you@yourdomain.com. We're the best alternative to Gmail, Microsoft 365, Proton Mail, Sendgrid, and Amazon SES – without hidden fees nor limits – and ultimately focused on 100% open-source software, quantum-resistant encryption, and privacy.

Unlike other services, we don't charge you per user. You get unlimited domains and aliases for only one monthly rate of $3/mo. All paid plans include 10 GB of SSD-backed encrypted SQLite storage (IMAP/POP3/CalDAV). Additional storage can be purchased for $3/mo per 10 GB of additional SSD-backed storage.

You can compare us to 56+ other email service providers on our Email Comparison page.

We provide email hosting and email forwarding service to 500,000+ domains and these notable users:

• Canonical
• Netflix
• The Linux Foundation
• The PHP Foundation
• Fox News Radio
• Disney Ad Sales
• jQuery
• LineageOS
• Ubuntu
• Kubuntu
• Lubuntu
• The University of Maryland
• The University of Washington
• Tufts University
• Swarthmore College
• Government of South Australia
• Government of Dominican Republic
• RCD Hotels
• International Correspondence Chess Federation
• John Graham-Cumming (Cloudflare)
• Isaac Z. Schlueter (npm)
• David Heinemeier Hansson (Ruby on Rails)

You can learn more about Forward Email on our About page.

Emails are delivered on average in under 5 seconds. We operate in real-time, unlike other providers which rely upon delayed queues.

We do not write to disk or store logs – with the exception of errors and outbound SMTP (see our Privacy Policy).

Everything is done in-memory and our source code is on GitHub.

1. If you have already registered your domain name somewhere, then you must completely skip this step and go to step two! Otherwise you can click here to register your domain name.
2. Do you remember where you registered your domain? Once you remember this, then follow the instructions below:
   Important: You must open a new tab and sign in to your domain registrar. You can easily click on your "Registrar" below to automatically do this. In this new tab, you must navigate to the DNS management page at your registrar – and we have provided the step by step navigation steps below under the "Steps to Configure" column. Once you've navigated to this page in the new tab, you can return to this tab and proceed to step three below. Do not close the opened tab yet; you will need it for future steps!

   Registrar	Steps to Configure
   1&1	Log in Domain Center (Select your domain) Edit DNS Settings
   Amazon Route 53	Log in Hosted Zones (Select your domain)
   Aplus.net	Log in My Servers Domain Management DNS Manager
   Bluehost	FOR ROCK: Log in Domains (Click the ▼ icon next to manage) DNS FOR LEGACY: Log in Domains Zone editor (Select your domain)
   Cloudflare	Log in DNS
   DNS Made Easy	Log in DNS (Select your domain)
   DNSimple	Log in (Select your domain) DNS Manage
   Digital Ocean	Log in Networking Domains (Select your domain) More Manage Domain
   Domain.com	Log in In card view, click manage on your domain In list view, click the gear icon DNS & Nameservers DNS Records
   Domains.com Watch	Log in (Select your domain) Manage (click gear icon) Click on DNS & Nameservers in left-hand menu
   DreamHost	Log in Panel Domains Manage Domains DNS
   Dyn	Log in Overview Manage Simple Editor Records
   Gandi	Log in (Select your domain) Management Edit the zone
   GoDaddy Watch	Log in Manage My Domains (Select your domain) Manage DNS
   Google Domains Watch	Log in (Select your domain) Configure DNS
   Namecheap Watch	Log in Domain List (Select your domain) Manage Advanced DNS
   Netlify	Log in (Select your domain) Setup Netlify DNS
   Network Solutions	Log in Account Manager My Domain Names (Select your domain) Manage Change Where Domain Points Advanced DNS
   Shopify Watch	Log in Managed Domains (Select your domain) DNS Settings
   Squarespace	Log in Home menu Settings Domains (Select your domain) Advanced settings Custom Records
   Vercel's Now	Using "now" CLI now dns add [domain] '@' MX [record-value] [priority]
   Weebly	Log in Domains page (Select your domain) DNS
   Wix	Log in Domains page (Click icon) Select Manage DNS Records
   eNom	Log in Domains My Domains
   Other	Important: Don't see your registrar name listed here? Simply search on the Internet for "how to change DNS records on $REGISTRAR" (replacing $REGISTRAR with the name of your registrar – e.g. "how to change DNS records on GoDaddy" if you're using GoDaddy).

3. Using your registrar's DNS management page (the other tab you have opened), set the following "MX" records:
   Important: Note that there should be NO other MX records set. Both records shown below MUST exist. Be sure there are no typos; and you have both mx1 and mx2 spelled correctly. If there were already MX records that existed, please delete them completely. The "TTL" value does not need to be 3600, it could be a lower or higher value if necessary.

   Name/Host/Alias	TTL	Type	Priority	Answer/Value
   "@", ".", or blank	3600	MX	10	mx1.forwardemail.net
   "@", ".", or blank	3600	MX	10	mx2.forwardemail.net

4. Using your registrar's DNS management page (the other tab you have opened), set the following TXT record(s):
   Important: If you are on a paid plan, then you must completely skip this step and go to step five! If you are not on a paid plan, then your forwarded addresses will be publicly searchable – go to My Account Domains and upgrade your domain to a paid plan if desired. If you would like to learn more about paid plans see our Pricing page. Otherwise you can continue to choose one or more combinations from Option A to Option F listed below.

   ---

   Option A: If you are forwarding all emails from your domain, (e.g. "all@sample.g0v.tw", "hello@sample.g0v.tw", etc) to a specific address "user@gmail.com":

   Name/Host/Alias	TTL	Type	Answer/Value
   "@", ".", or blank	3600	TXT	forward-email=user@gmail.com Copy Encrypt

   Tip: Make sure to replace the values above in the "Value" column with your own email address. The "TTL" value does not need to be 3600, it could be a lower or higher value if necessary. A lower time to live ("TTL") value will ensure any future changes made to your DNS records are propagated throughout the Internet quicker – think of this as how long it will be cached in-memory (in seconds). You can learn more about TTL on Wikipedia.

   ---

   Option B: If you just need to forward a single email address (e.g. hello@sample.g0v.tw to user@gmail.com; this will also forward "hello+test@sample.g0v.tw" to "user+test@gmail.com" automatically):

   Name/Host/Alias	TTL	Type	Answer/Value
   "@", ".", or blank	3600	TXT	forward-email=hello:user@gmail.com Copy Encrypt

   ---

   Option C: If you are forwarding multiple emails, then you'll want to separate them with a comma:

   Name/Host/Alias	TTL	Type	Answer/Value
   "@", ".", or blank	3600	TXT	forward-email=hello:user@gmail.com,support:user@gmail.com Copy Encrypt

   ---

   Option D: You can have an infinite amount of forwarding emails setup – just make sure to not wrap over 255 characters in a single-line and start each line with "forward-email=". An example is provided below:

   Name/Host/Alias	TTL	Type	Answer/Value
   "@", ".", or blank	3600	TXT	forward-email=hello:user@gmail.com,support:user@gmail.com Copy Encrypt
   "@", ".", or blank	3600	TXT	forward-email=help:user@gmail.com,foo:user@gmail.com Copy Encrypt
   "@", ".", or blank	3600	TXT	forward-email=orders:user@gmail.com,baz:user@gmail.com Copy Encrypt
   "@", ".", or blank	3600	TXT	forward-email=info:user@gmail.com,beep:user@gmail.com Copy Encrypt
   "@", ".", or blank	3600	TXT	forward-email=errors:user@gmail.com,boop:user@gmail.com Copy Encrypt

   ---

   Option E: You can also specify a domain name in your TXT record to have global alias forwarding (e.g. "user@sample.g0v.tw" will get forwarded to "user@example.net"):

   Name/Host/Alias	TTL	Type	Answer/Value
   "@", ".", or blank	3600	TXT	forward-email=example.net Copy Encrypt

   ---

   Option F: You can even use webhooks as a global or individual alias to forward emails to. See the example and full section on webhooks titled Do you support webhooks below.

   Name/Host/Alias	TTL	Type	Answer/Value
   "@", ".", or blank	3600	TXT	forward-email=alias:https://requestbin.com/r/en8pfhdgcculn Copy Encrypt

   ---

   Option G: You can even use regular expressions ("regex") for matching aliases and for handling substitutions to forward emails to. See the examples and full section on regex titled Do you support regular expressions or regex below.
   Need advanced regex with substitution? See the examples and full section on regex titled Do you support regular expressions or regex below.
   Simple Example: If I want all emails that go to `linus@sample.g0v.tw` or `torvalds@sample.g0v.tw` to forward to `user@gmail.com`:

   Name/Host/Alias	TTL	Type	Answer/Value
   "@", ".", or blank	3600	TXT	forward-email=/^(linus|torvalds)$/:user@gmail.com Copy Encrypt

   ---

5. Using your registrar's DNS management page (the other tab you have opened), additionally set the following TXT record:

   Name/Host/Alias	TTL	Type	Answer/Value
   "@", ".", or blank	3600	TXT	v=spf1 a include:spf.forwardemail.net -all

   Important: If you are using Gmail (e.g. Send Mail As) or G Suite, then you'll need to append include:_spf.google.com to the value above, for example:
   
   v=spf1 a include:spf.forwardemail.net include:_spf.google.com -all
   Tip: If you already have a similar line with "v=spf1", then you'll need to append include:spf.forwardemail.net right before any existing "include:host.com" records and before the "-all" in the same line, for example:
   
   v=spf1 a include:spf.forwardemail.net include:host.com -all
   
   Note that there is a difference between "-all" and "~all". The "-" indicates that the SPF check should FAIL if it does not match, and "~" indicates that the SPF check should SOFTFAIL. We recommend to use the "-all" approach to prevent domain forgery.
   
   You may also need to include the SPF record for whichever host you are sending mail from (e.g. Outlook).
6. Verify your DNS records using our "Verify Records" tool available at My Account Domains Setup.
7. Send a test email to confirm it works. Note that it might take some time for your DNS records to propagate.
   Tip: If you are not receiving test emails, or receive a test email that says "Be careful with this message", then see the answers for Why am I not receiving my test emails and Why are my test emails sent to myself in Gmail showing as "suspicious" respectively.
8. If you wish to "Send Mail As" from Gmail, then you will need to watch this video, or follow the steps under How to Send Mail As Using Gmail below.

Yes, but you should only have one MX exchange listed in your DNS records.

Do not attempt to use "Priority" as a way to configure multiple MX exchanges.

Instead, you need to configure your existing MX exchange to forward mail for all non-matching aliases to our service's exchanges (mx1.forwardemail.net and/or mx2.forwardemail.net).

If you are using Google Workspace and you want to forward all non-matching aliases to our service, then see https://support.google.com/a/answer/6297084.

If you are using Microsoft 365 (Outlook) and you want to forward all non-matching aliases to our service, then see https://learn.microsoft.com/en-us/exchange/mail-flow-best-practices/use-connectors-to-configure-mail-flow/set-up-connectors-to-route-mail and https://learn.microsoft.com/en-us/exchange/mail-flow-best-practices/manage-mail-flow-for-multiple-locations.

Using your registrar's DNS management page, set the following TXT record:

Go to My Account Domains Settings Outbound SMTP Configuration and follow setup instructions.

Go to My Account Domains Settings Outbound SMTP Configuration and follow setup instructions.

We automatically configure Sender Rewriting Scheme ("SRS") – you do not need to do this yourself.

Please refer to our section on MTA-STS for more insight.

Yes, as of May 2023 we support sending email with SMTP as an add-on for all paid users.

We rate limit users and domains to 300 outbound SMTP messages per 1 day. This averages 9000+ emails in a calendar month. If you need to exceed this amount or have consistently large emails, then please contact us.

Our server is smtp.forwardemail.net and is also monitored on our status page.

It supports both IPv4 and IPv6 and is available over ports 465 and 2465 for SSL/TLS and 587, 2587, 2525, and 25 for TLS (STARTTLS).

In order to send outbound email with SMTP, the SMTP user must be the email address of an alias that exists for the domain at My Account Domains – and the SMTP password must be an alias-specific generated password.

Please refer to Do you support sending email with SMTP for step by step instructions.

Our server is imap.forwardemail.net and is also monitored on our status page.

It supports both IPv4 and IPv6 and is available over ports 993 and 2993 for SSL/TLS.

In order to connect with IMAP, the IMAP user must be the email address of an alias that exists for the domain at My Account Domains – and the IMAP password must be an alias-specific generated password.

Please refer to Do you support receiving email with IMAP for step by step instructions.

Our server is pop3.forwardemail.net and is also monitored on our status page.

It supports both IPv4 and IPv6 and is available over ports 995 and 2995 for SSL/TLS.

In order to connect with POP3, the POP3 user must be the email address of an alias that exists for the domain at My Account Domains – and the IMAP password must be an alias-specific generated password.

Please refer to Do you support POP3 for step by step instructions.

Yes, as of May 2023 we support sending email with API as an add-on for all paid users.

Please view our section on Emails in our API documentation for options, examples, and more insight.

In order to send outbound email with our API, you must use your API token available under My Security.

Yes, as of October 16, 2023 we support receiving email over IMAP as an add-on for all paid users. Please read our deep-dive article on how our encrypted SQLite mailbox storage feature works.

Yes, we support OpenPGP, end-to-end encryption ("E2EE"), and the discovery of public keys using Web Key Directory ("WKD"). You can configure OpenPGP using keys.openpgp.org or self-host your own keys (refer to this gist for WKD server setup).

1. Download and install your email client's recommended plugin below:

   Email Client	Platform	Recommended Plugin	Notes
   Thunderbird	Desktop	Configure OpenPGP in Thunderbird	Thunderbird has built-in support for OpenPGP.
   Gmail	Browser	Mailvelope or FlowCrypt (proprietary license)	Gmail does not support OpenPGP, however you can download the open-source plugin Mailvelope or FlowCrypt.
   Apple Mail	macOS	Free-GPGMail	Apple Mail does not support OpenPGP, however you can download the open-source plugin Free-GPGMail.
   Apple Mail	iOS	PGPro or FlowCrypt (proprietary license)	Apple Mail does not support OpenPGP, however you can download the open-source plugin PGPro or FlowCrypt.
   Outlook	Windows	gpg4win	Outlook's desktop mail client does not support OpenPGP, however you can download the open-source plugin gpg4win.
   Outlook	Browser	Mailvelope or FlowCrypt (proprietary license)	Outlook's web-based mail client does not support OpenPGP, however you can download the open-source plugin Mailvelope or FlowCrypt.
   Android	Mobile	OpenKeychain or FlowCrypt	Android mail clients such as K-9 Mail and FairEmail both support the open-source plugin OpenKeychain. You could alternatively use the open-source (proprietary licensing) plugin FlowCrypt.
   Google Chrome	Browser	Mailvelope or FlowCrypt (proprietary license)	You can download the open-source browser extension Mailvelope or FlowCrypt.
   Mozilla Firefox	Browser	Mailvelope or FlowCrypt (proprietary license)	You can download the open-source browser extension Mailvelope or FlowCrypt.
   Microsoft Edge	Browser	Mailvelope	You can download the open-source browser extension Mailvelope.
   Brave	Browser	Mailvelope or FlowCrypt (proprietary license)	You can download the open-source browser extension Mailvelope or FlowCrypt.
   Balsa	Desktop	Configure OpenPGP in Balsa	Balsa has built-in support for OpenPGP.
   KMail	Desktop	Configure OpenPGP in KMail	KMail has built-in support for OpenPGP.
   GNOME Evolution	Desktop	Configure OpenPGP in Evolution	GNOME Evolution has built-in support for OpenPGP.
   Terminal	Desktop	Configure gpg in Terminal	You can use the open-source gpg command line tool to generate a new key from command line.

2. Open the plugin, create your public key, and configure your email client to use it.

3. Upload your public key at https://keys.openpgp.org/upload.

   Tip: You can visit https://keys.openpgp.org/manage to manage your key in the future.
   Optional Add-on: If you are using our encrypted storage (IMAP/POP3) service and want all email stored in your (already encrypted) SQLite database to be encrypted with your public key, then go to My Account Domains Aliases (e.g. hello@sample.g0v.tw) Edit OpenPGP and upload your public key.

4. Add a new CNAME record to your domain name (e.g. sample.g0v.tw):

   Name/Host/Alias	TTL	Type	Answer/Value
   openpgpkey	3600	CNAME	wkd.keys.openpgp.org

   Tip: If your alias is using our vanity/disposable domains (e.g. hideaddress.net), then you can skip this step.

Yes, as of December 4, 2023 we support POP3 as an add-on for all paid users. Please read our deep-dive article on how our encrypted SQLite mailbox storage feature works.

Yes, as of February 5, 2024 we have added this feature. Our server is caldav.forwardemail.net and is also monitored on our status page.

It supports both IPv4 and IPv6 and is available over port 443 (HTTPS).

In order to use calendar support, the user must be the email address of an alias that exists for the domain at My Account Domains – and the password must be an alias-specific generated password.

How do I connect and configure my calendars

To configure your calendar, use the CalDAV URL of: https://caldav.forwardemail.net (or simply caldav.forwardemail.net if your client allows it)

How do I add more calendars and manage existing calendars

If you'd like to add additional calendars, then just add a new calendar URL of: https://caldav.forwardemail.net/dav/principals/calendar-name (be sure to replace calendar-name with your desired calendar name)

You can change a calendar's name and color after creation – just use your preferred calendar application (e.g. Apple Mail or Thunderbird).

If you're sending a test email to yourself, then it may not show up in your inbox because it has the same "Message-ID" header.

This is a widely known issue, and also affects services such as Gmail. Here is the official Gmail answer regarding this issue.

If you continue to have issues, then it is most likely to be an issue with DNS propagation. You will need to wait a bit longer and try again (or try setting a lower TTL value on your TXT records).

Still having issues? Please contact us so we can help investigate the issue and find a quick resolution.

At anytime you can export your mailboxes as EML, Mbox, or encrypted SQLite formats.

Go to My Account Domains Aliases Download Backup and select your preferred export format type.

You will be emailed a link to download the export once it has finished.

Note that this export download link expires after 4 hours for security concerns.

If you need to inspect your exported EML or Mbox formats, then these open-soruce tools may be useful:

Additionally if you need to convert a Mbox file to EML file, then you can use https://github.com/noelmartinon/mboxzilla.

You can easily import your email to Forward Email (e.g. using Thunderbird) with the instructions below:

1. Export your email from your existing email provider:

   Email Provider	Export Format	Export Instructions
   Gmail	MBOX	https://takeout.google.com/settings/takeout/custom/gmail
   Outlook	PST	Tip: If you are using Outlook (PST export format), then you could simply follow the instructions under "Other" below. However we have provided links below to convert PST to MBOX/EML format based off your operating system: Zinkuba for Windows (GitHub) readpst for Windows cygwin – (e.g. readpst -u -o $OUT_DIR $IN_DIR replacing $OUT_DIR and $IN_DIR with the output directory and input directory paths respectively). readpst for Ubuntu/Linux – (e.g. sudo apt-get install readpst and then readpst -u -o $OUT_DIR $IN_DIR, replacing $OUT_DIR and $IN_DIR with the output directory and input directory paths respectively). readpst for macOS (via brew) – (e.g. brew install libpst and then readpst -u -o $OUT_DIR $IN_DIR, replacing $OUT_DIR and $IN_DIR with the output directory and input directory paths respectively). PST Converter for Windows (GitHub)
   Apple Mail	MBOX	https://support.apple.com/guide/mail/import-or-export-mailboxes-mlhlp1030/mac#apd37a3190755974
   Fastmail	EML	https://www.fastmail.help/hc/en-us/articles/360060590573-Download-all-your-data#downloadmail
   Proton Mail	MBOX/EML	https://proton.me/support/export-emails-import-export-app
   Tutanota	EML	https://github.com/crepererum-oss/tatuanatata?tab=readme-ov-file#tatutanatata
   Gandi	EML	https://docs.gandi.net/en/gandimail/common_operations/backup_email.html#contents
   Zoho	EML	https://www.zoho.com/mail/help/import-export-emails.html#alink2
   Other	Use Thunderbird	Set up your existing email account in Thunderbird and then use the ImportExportTools NG plugin to export and import your email. You may also be able to simply copy/paste or drag/drop emails between one account to another.

2. Download, install, and open Thunderbird.

3. Create a new account using your alias' full email address (e.g. you@yourdomain.com) and your generated password. If you do not yet have a generated password, then refer to our setup instructions.

4. Download and install the ImportExportTools NG Thunderbird plugin.

5. Create a new local folder in Thunderbird, and then right click on it → select the ImportExportTools NG option → choose Import mbox file (for MBOX export format) – or – Import messages / Import all messages from a directory (for EML export format).

6. Drag/drop from the local folder to a new (or existing) IMAP folder in Thunderbird you wish to upload messages to in IMAP storage with our service. This will ensure they are backed up online with our SQLite encrypted storage.

   Tip: If you are confused as to how to import into Thunderbird, then you can refer to official instructions at https://kb.mozillazine.org/Importing_folders and https://github.com/thunderbird/import-export-tools-ng/wiki.

Email relies on the SMTP protocol. This protocol consists of commands sent to a server (running most commonly on port 25). There is an initial connection, then the sender indicates who the mail is from ("MAIL FROM"), followed by where it's going to ("RCPT TO"), and finally the headers and the body of the email itself ("DATA"). The flow of our email forwarding system is described relative to each SMTP protocol command below:

• Initial Connection (no command name, e.g. telnet sample.g0v.tw 25) - This is the initial connection. We check senders that aren't in our allowlist against our denylist. Finally, if a sender is not in our allowlist, then we check to see if they have been greylisted.

• HELO - This indicates a greeting to identify the sender's FQDN, IP address, or mail handler name. This value can be spoofed, so we do not rely on this data and instead use the reverse hostname lookup of the connection's IP address.

• MAIL FROM - This indicates the envelope mail from address of the email. If a value is entered, it must be a valid RFC 5322 email address. Empty values are permitted. We check for backscatter here, and we also check the MAIL FROM against our denylist. We finally check senders that are not on the allowlist for rate limiting (see the section on Rate Limiting and allowlist for more information).

• RCPT TO - This indicates the recipient(s) of the email. These must be valid RFC 5322 email addresses. We only permit up to 50 envelope recipients per message (this is different than the "To" header from an email). We also check for a valid Sender Rewriting Scheme ("SRS") address here to protect against spoofing with our SRS domain name. We also check the recipient against our denylist.

• DATA - This is the core part of our service which processes an email. See the section How do you process an email for forwarding below for more insight.

This section describes our process related to the SMTP protocol command DATA in the section How does your email forwarding system work above – it is how we process an email's headers, body, security, determine where it needs to be delivered to, and how we handle connections.

1. If the message exceeds the maximum size of 50mb, then it is rejected with a 552 error code.

2. If the message did not have any headers, or headers were unable to be parsed, then it is rejected with a 421 error code.

3. If the message did not contain a "From" header, or if any of the values in the "From" header were not valid RFC 5322 email addresses, then it is rejected with a 550 error code.

4. If the message had more than 25 "Received" headers, then it was determined to have been stuck in a redirect loop, and it is rejected with a 550 error code.

5. Using the email's fingerprint (see the section on Fingerprinting), we will check to see the message has been attempted to be retried for more than 5 days (which matches default postfix behavior), and if so, then it will be rejected with a 550 error code.

6. If the message had a "To" header, and if any of the message's "To" headers were directed towards SRS rewritten addresses using our domain name, then we will rewrite them (this namely applies for vacation responders).

7. We store in-memory the results from scanning the email using Spam Scanner.

8. If there were any arbitrary results from Spam Scanner, then it is rejected with a 554 error code. Arbitrary results only include the GTUBE test at the time of this writing. See https://spamassassin.apache.org/gtube/ for more insight.

9. We will add the following headers to the message for debugging and abuse prevention purposes:

   • X-Original-To - the original RCPT TO email address for the message.
     • This header's value has Bcc header parsed addresses removed from it.
     • This is useful for determining where an email was originally delivered to.
   • X-ForwardEmail-Version - the current SemVer version from package.json of our codebase.
   • X-ForwardEmail-Session-ID - a session ID value used for debug purposes (only applies in non-production environments).
   • X-ForwardEmail-Sender - a comma separated list containing the original envelope MAIL FROM address (if it was not blank), the reverse PTR client FQDN (if it exists), and the sender's IP address.
   • X-ForwardEmail-ID - this is only applicable for outbound SMTP and correlates to the email ID stored in My Account → Emails
   • X-Report-Abuse - with a value of abuse@forwardemail.net.
   • X-Report-Abuse-To - with a value of abuse@forwardemail.net.
   • X-Complaints-To - with a value of abuse@forwardemail.net.

10. We then check the message for DKIM, SPF, ARC, and DMARC.

    • If the message failed DMARC and the domain had a rejection policy (e.g. p=reject was in the DMARC policy), then it is rejected with a 550 error code. Typically a DMARC policy for a domain can be found in the _dmarc sub-domain TXT record, (e.g. dig _dmarc.sample.g0v.tw txt).
    • If the message failed SPF and the domain had a hard fail policy (e.g. -all was in the SPF policy as opposed to ~all or no policy at all), then it is rejected with a 550 error code. Typically an SPF policy for a domain can be found in the TXT record for the root domain (e.g. dig sample.g0v.tw txt). See this section for more information on sending mail as with Gmail regarding SPF.

11. Now we process the recipients of the message as collected from the RCPT TO command in the section How does your email forwarding system work above. For each recipient, we perform the following operations:

    • We lookup the TXT records of the domain name (the part after the @ symbol, e.g. sample.g0v.tw if the email address was test@sample.g0v.tw). For example, if the domain is sample.g0v.tw we do a DNS lookup such as dig sample.g0v.tw txt.
    • We parse all TXT records that start with either forward-email= (free plans) or forward-email-site-verification= (paid plans). Note that we parse both, in order to process emails while a user is upgrading or downgrading plans.
    • From these parsed TXT records, we iterate over them to extract the forwarding configuration (as described in the section How do I get started and set up email forwarding above). Note that we only support one forward-email-site-verification= value, and if more than one is supplied, then a 550 error will occur and the sender will receive a bounce for this recipient.
    • Recursively we iterate over the extracted forwarding configuration to determine global forwarding, regex based forwarding, and all other supported forwarding configurations – which are now known as our "Forwarding Addresses".
    • For each Forwarding Address, we support one recursive lookup (which will start this series of operations over on the given address). If a recursive match was found, then the parent result will be removed from Forwarding Addresses, and the children added.
    • Forwarding Addresses are parsed for uniqueness (since we don't want to send duplicates to one address or spawn additionally unnecessary SMTP client connections).
    • For each Forwarding Address, we lookup its domain name against our API endpoint /v1/max-forwarded-addresses (in order to determine how many addresses the domain is permitted to forward email to per alias, e.g. 10 by default – see the section on maximum limit on forwarding per alias). If this limit is exceeded, then a 550 error will occur and the sender will receive a bounce for this recipient.
    • We lookup the settings of the original recipient against our API endpoint /v1/settings, which supports a lookup for paid users (with a fallback for free users). This returns a configuration object for advanced settings for port (Number, e.g. 25), has_adult_content_protection (Boolean), has_phishing_protection (Boolean), has_executable_protection (Boolean), and has_virus_protection (Boolean).
    • Based off these settings, we then check against Spam Scanner results and if any errors occur, then the message is rejected with a 554 error code (e.g. if has_virus_protection is enabled, then we will check the Spam Scanner results for viruses). Note that all free plan users will be opted-in for checks against adult-content, phishing, executables, and viruses. By default, all paid plan users are opted-in as well, but this configuration can be altered under the Settings page for a domain in the Forward Email dashboard).

12. For each processed recipient's Forwarding Addresses, we then perform the following operations:

    • The address is checked against our denylist, and if it was listed, then a 421 error code will occur (indicates to sender to retry again later).
    • If the address is a webhook, then we set a Boolean for future operations (see below – we group together similar webhooks to make one POST request vs. multiple for delivery).
    • If the address is an email address, then we parse the host for future operations (see below – we group together similar hosts to make one connection vs. multiple individual connections for delivery).

13. If there are no recipients and there are no bounces, then we respond with a 550 error of "Invalid recipients".

14. If there are recipients, then we iterate over them (grouped together by the same host) and deliver the emails. See the section How do you handle email delivery issues below for more insight.

    • If any errors occur while sending emails, then we will store them in-memory for later processing.
    • We will take the lowest error code (if any) from sending emails – and use that as the response code to the DATA command. This means that emails not delivered will typically be retried by the original sender, yet emails that were already delivered will not be re-sent the next time the message is sent (as we use Fingerprinting).
    • If no errors occurred, then we will send a 250 successful SMTP response status code.
    • A bounce is determined to be any delivery attempted that results in a status code that is >= 500 (permanent failures).

15. If no bounces occurred (permanent failures), then we will return a SMTP response status code of the lowest error code from non-permanent failures (or a 250 successful status code if there were none).

16. If bounces did occur then we will send bounce emails in the background after returning the lowest of all error codes to the sender. However, if the lowest error code is >= 500, then we do not send any bounce emails. This is because if we did, then senders would receive a double bounce email (e.g. one from their outbound MTA, such as Gmail – and also one from us). See the section on How do you protect against backscatter below for more insight.

Note that we will do a "Friendly-From" rewrite on the emails if and only if the DMARC policy of the sender was p=reject, AND it has a passing SPF, AND no DKIM signatures were aligned with the "From" header. This means that we will alter the "From" header on the message, set "X-Original-From", and also set a "Reply-To" if it was not already set. We will also re-seal the ARC seal on the message after altering these headers.

We also use smart-parsing of error messages at every level of our stack – in our code, DNS requests, Node.js internals, HTTP requests (e.g. 408, 413, and 429 are mapped to the SMTP response code of 421 if the recipient is a webhook), and mail server responses (e.g. responses with "defer" or "slowdown" would be retried as 421 errors).

Our logic is dummy-proof and it will also retry for SSL/TLS errors, connection issues, and more. The goal with dummy-proofing is to maximize deliverability to all recipients for a forwarding configuration.

If the recipient is a webhook, then we will permit a 60 second timeout for the request to complete with up to 3 retries (so 4 requests total before a failure). Note that we correctly parse error codes 408, 413, and 429 and map them to a SMTP response code of 421.

Otherwise if the recipient is an email address, then we will attempt to send the email with opportunistic TLS (we attempt to use STARTTLS if it is available on the recipient mail server). If a SSL/TLS error occurs while attempting to send the email, then we will attempt to send the email without TLS (without using STARTTLS).

If any DNS or connection errors occur, then we will return to the DATA command a SMTP response code of 421, otherwise if there are >= 500 level errors, then bounces will be sent.

If we detect that an email server we are attempting to deliver to has one or more of our mail exchange IP addresses blocked (e.g. by whatever technology they use for deferring spammers), then we will send a SMTP response code of 421 for the sender to retry their message later (and we are alerted to the issue so we can hopefully resolve it before the next attempt).

We routinely monitor all major DNS denylists and if any of our mail exchange ("MX") IP addresses are listed in a major denylist, we will pull it out of the relevant DNS A record round robin if possible until it the issue is resolved.

At the time of this writing, we are listed in several DNS allowlists as well, and we take monitoring denylists seriously. If you see any issues before we have a chance to resolve them, please notify us in writing at support@forwardemail.net.

Our IP addresses are publicly available, see this section below for more insight.

Email usernames equal to any of the following (case-insensitive) are considered to be no-reply addresses:

• do-not-reply
• do-not-respond
• do.not.reply
• donotreply
• donotrespond
• dont-reply
• naoresponda
• no-replies
• no-reply
• no-replys
• no.replies
• no.reply
• no.replys
• no_reply
• nobody
• noreplies
• noreply
• noreplys

This list is maintained as an open-source project on GitHub.

We publish our IP addresses at https://forwardemail.net/ips.

Yes, we have a list of domain name extensions that are allowlisted by default and a dynamic, cached, and rolling allowlist based off strict criteria.

All emails, domains, and recipients from customers on paid plans are automatically added to our allowlist.

What domain name extensions are allowlisted by default

The following domain name extensions are considered to be allowlisted by default (regardless if they are on the Umbrella Popularity List or not):

• edu
• gov
• mil
• int
• arpa
• dni.us
• fed.us
• isa.us
• kids.us
• nsn.us
• ak.us
• al.us
• ar.us
• as.us
• az.us
• ca.us
• co.us
• ct.us
• dc.us
• de.us
• fl.us
• ga.us
• gu.us
• hi.us
• ia.us
• id.us
• il.us
• in.us
• ks.us
• ky.us
• la.us
• ma.us
• md.us
• me.us
• mi.us
• mn.us
• mo.us
• ms.us
• mt.us
• nc.us
• nd.us
• ne.us
• nh.us
• nj.us
• nm.us
• nv.us
• ny.us
• oh.us
• ok.us
• or.us
• pa.us
• pr.us
• ri.us
• sc.us
• sd.us
• tn.us
• tx.us
• ut.us
• va.us
• vi.us
• vt.us
• wa.us
• wi.us
• wv.us
• wy.us
• mil.tt
• edu.tt
• edu.tr
• edu.ua
• edu.au
• ac.at
• edu.br
• ac.nz
• school.nz
• cri.nz
• health.nz
• mil.nz
• parliament.nz
• ac.in
• edu.in
• mil.in
• ac.jp
• ed.jp
• lg.jp
• ac.za
• edu.za
• mil.za
• school.za
• mil.kr
• ac.kr
• hs.kr
• ms.kr
• es.kr
• sc.kr
• kg.kr
• edu.es
• ac.lk
• sch.lk
• edu.lk
• ac.th
• mi.th
• admin.ch
• canada.ca
• gc.ca
• go.id
• go.jp
• go.ke
• go.kr
• go.th
• gob.ar
• gob.cl
• gob.es
• gob.mx
• gob.ve
• gob.sv
• gouv.fr
• gouv.nc
• gouv.qc.ca
• gov.ad
• gov.af
• gov.ai
• gov.al
• gov.am
• gov.ao
• gov.au
• gov.aw
• gov.ax
• gov.az
• gov.bd
• gov.be
• gov.bg
• gov.bm
• gov.by
• gov.cl
• gov.cn
• gov.co
• gov.cy
• gov.cz
• gov.dz
• gov.eg
• gov.fi
• gov.fk
• gov.gg
• gov.gr
• gov.hk
• gov.hr
• gov.hu
• gov.ie
• gov.il
• gov.im
• gov.in
• gov.iq
• gov.ir
• gov.it
• gov.je
• gov.kp
• gov.krd
• gov.ky
• gov.kz
• gov.lb
• gov.lk
• gov.lt
• gov.lv
• gov.ma
• gov.mm
• gov.mo
• gov.mt
• gov.my
• gov.ng
• gov.np
• gov.ph
• gov.pk
• gov.pl
• gov.pt
• gov.py
• gov.ro
• gov.ru
• gov.scot
• gov.se
• gov.sg
• gov.si
• gov.sk
• gov.tr
• gov.tt
• gov.tw
• gov.ua
• gov.uk
• gov.vn
• gov.wales
• gov.za
• government.pn
• govt.nz
• gv.at
• ac.uk
• bl.uk
• judiciary.uk
• mod.uk
• nhs.uk
• parliament.uk
• police.uk
• rct.uk
• royal.uk
• sch.uk
• ukaea.uk

Additionally these brand and corporate top-level domains are allowlisted by default (e.g. apple for applecard.apple for Apple Card bank statements):

• aaa
• aarp
• abarth
• abb
• abbott
• abbvie
• abc
• accenture
• aco
• aeg
• aetna
• afl
• agakhan
• aig
• aigo
• airbus
• airtel
• akdn
• alfaromeo
• alibaba
• alipay
• allfinanz
• allstate
• ally
• alstom
• amazon
• americanexpress
• amex
• amica
• android
• anz
• aol
• apple
• aquarelle
• aramco
• audi
• auspost
• aws
• axa
• azure
• baidu
• bananarepublic
• barclaycard
• barclays
• basketball
• bauhaus
• bbc
• bbt
• bbva
• bcg
• bentley
• bharti
• bing
• blanco
• bloomberg
• bms
• bmw
• bnl
• bnpparibas
• boehringer
• bond
• booking
• bosch
• bostik
• bradesco
• bridgestone
• brother
• bugatti
• cal
• calvinklein
• canon
• capitalone
• caravan
• cartier
• cba
• cbn
• cbre
• cbs
• cern
• cfa
• chanel
• chase
• chintai
• chrome
• chrysler
• cipriani
• cisco
• citadel
• citi
• citic
• clubmed
• comcast
• commbank
• creditunion
• crown
• crs
• csc
• cuisinella
• dabur
• datsun
• dealer
• dell
• deloitte
• delta
• dhl
• discover
• dish
• dnp
• dodge
• dunlop
• dupont
• dvag
• edeka
• emerck
• epson
• ericsson
• erni
• esurance
• etisalat
• eurovision
• everbank
• extraspace
• fage
• fairwinds
• farmers
• fedex
• ferrari
• ferrero
• fiat
• fidelity
• firestone
• firmdale
• flickr
• flir
• flsmidth
• ford
• fox
• fresenius
• forex
• frogans
• frontier
• fujitsu
• fujixerox
• gallo
• gallup
• gap
• gbiz
• gea
• genting
• giving
• gle
• globo
• gmail
• gmo
• gmx
• godaddy
• goldpoint
• goodyear
• goog
• google
• grainger
• guardian
• gucci
• hbo
• hdfc
• hdfcbank
• hermes
• hisamitsu
• hitachi
• hkt
• honda
• honeywell
• hotmail
• hsbc
• hughes
• hyatt
• hyundai
• ibm
• ieee
• ifm
• ikano
• imdb
• infiniti
• intel
• intuit
• ipiranga
• iselect
• itau
• itv
• iveco
• jaguar
• java
• jcb
• jcp
• jeep
• jpmorgan
• juniper
• kddi
• kerryhotels
• kerrylogistics
• kerryproperties
• kfh
• kia
• kinder
• kindle
• komatsu
• kpmg
• kred
• kuokgroup
• lacaixa
• ladbrokes
• lamborghini
• lancaster
• lancia
• lancome
• landrover
• lanxess
• lasalle
• latrobe
• lds
• leclerc
• lego
• liaison
• lexus
• lidl
• lifestyle
• lilly
• lincoln
• linde
• lipsy
• lixil
• locus
• lotte
• lpl
• lplfinancial
• lundbeck
• lupin
• macys
• maif
• man
• mango
• marriott
• maserati
• mattel
• mckinsey
• metlife
• microsoft
• mini
• mit
• mitsubishi
• mlb
• mma
• monash
• mormon
• moto
• movistar
• msd
• mtn
• mtr
• mutual
• nadex
• nationwide
• natura
• nba
• nec
• netflix
• neustar
• newholland
• nfl
• nhk
• nico
• nike
• nikon
• nissan
• nissay
• nokia
• northwesternmutual
• norton
• nra
• ntt
• obi
• office
• omega
• oracle
• orange
• otsuka
• ovh
• panasonic
• pccw
• pfizer
• philips
• piaget
• pictet
• ping
• pioneer
• play
• playstation
• pohl
• politie
• praxi
• prod
• progressive
• pru
• prudential
• pwc
• quest
• qvc
• redstone
• reliance
• rexroth
• ricoh
• rmit
• rocher
• rogers
• rwe
• safety
• sakura
• samsung
• sandvik
• sandvikcoromant
• sanofi
• sap
• saxo
• sbi
• sbs
• sca
• scb
• schaeffler
• schmidt
• schwarz
• scjohnson
• scor
• seat
• sener
• ses
• sew
• seven
• sfr
• seek
• shangrila
• sharp
• shaw
• shell
• shriram
• sina
• sky
• skype
• smart
• sncf
• softbank
• sohu
• sony
• spiegel
• stada
• staples
• star
• starhub
• statebank
• statefarm
• statoil
• stc
• stcgroup
• suzuki
• swatch
• swiftcover
• symantec
• taobao
• target
• tatamotors
• tdk
• telecity
• telefonica
• temasek
• teva
• tiffany
• tjx
• toray
• toshiba
• total
• toyota
• travelchannel
• travelers
• tui
• tvs
• ubs
• unicom
• uol
• ups
• vanguard
• verisign
• vig
• viking
• virgin
• visa
• vista
• vistaprint
• vivo
• volkswagen
• volvo
• walmart
• walter
• weatherchannel
• weber
• weir
• williamhill
• windows
• wme
• wolterskluwer
• woodside
• wtc
• xbox
• xerox
• xfinity
• yahoo
• yamaxun
• yandex
• yodobashi
• youtube
• zappos
• zara
• zippo

What is your allowlist criteria

We have a static list of domain name extensions allowlisted by default – and we also maintain a dynamic, cached, rolling allowlist based off the following strict criteria:

• Sender root domain must be of a domain name extension that matches the list we offer on our free plan (with the addition of biz and info). We also include edu, gov, and mil partial matches, such as xyz.gov.au and xyz.edu.au.
• Sender root domain must be within top 100,000 unique root domain parsed results from Umbrella Popularity List ("UPL").
• Sender root domain must be within top 50,000 results from unique root domains appearing in at least 4 of past 7 days of UPL's (~50%+).
• Sender root domain must not be categorized as adult-content or malware by Cloudflare.
• Sender root domain must have either A or MX records set.
• Sender root domain must have either A record(s), MX record(s), DMARC record with p=reject or p=quarantine, or an SPF record with -all or ~all qualifier.

If this criteria is satisfied, then the sender root domain will be cached for 7 days. Note that our automated job runs daily – therefore this is a rolling allowlist cache that updates daily.

Our automated job will download the previous 7 days of UPL's in-memory, unzip them, and then parse in-memory according to the strict criteria above.

Popular domains at the time of this writing such as Google, Yahoo, Microsoft, Amazon, Meta, Twitter, Netflix, Spotify, and more – are of course included.

If you are a sender not in our allowlist, then the first time your FQDN root domain or IP address sends an email, you will be rate limited and greylisted. Note that this is standard practice adopted as an email standard. Most email server clients will attempt to retry if they receive a rate limit or greylist error (e.g. a 421 or 4xx level error status code).

Note that specific senders such as a@gmail.com, b@xyz.edu, and c@gov.au can still be denylisted (e.g. if we automatically detect spam, phishing, or malware from those senders).

As of March 31, 2023 we enforced a new blanket spam rule to protect our users and service.

This new rule allows only the following domain name extensions to be used on our free plan:

• ac
• ad
• ag
• ai
• al
• am
• app
• as
• at
• au
• ba
• be
• br
• by
• ca
• cc
• cd
• ch
• ck
• co
• com
• de
• dev
• dj
• dk
• ee
• es
• eu
• family
• fi
• fm
• fr
• gg
• gl
• id
• ie
• il
• im
• in
• io
• ir
• is
• it
• je
• jp
• ke
• kr
• la
• li
• lv
• ly
• md
• me
• mn
• ms
• mu
• mx
• net
• ni
• nl
• no
• nu
• nz
• org
• pl
• pr
• pt
• pw
• rs
• sc
• se
• sh
• si
• sm
• sr
• st
• tc
• tm
• to
• tv
• uk
• us
• uz
• vc
• vg
• vu
• ws
• xyz
• za

Yes, we have a very lax email greylisting policy used. Greylisting only applies for senders not on our allowlist and lasts in our cache for 30 days.

For any new sender, we store a key in our Redis database for 30 days with a value set to the initial arrival time of their first request. We then reject their email with a retry status code of 450 and only allow it to pass once 5 minutes has passed.

If they have successfully waited for 5 minutes from this initial arrival time, then their emails will be accepted and they will not receive this 450 status code.

The key consists of either the FQDN root domain or the sender's IP address. This means that any sub-domain that passes the greylist also will pass for the root domain, and vice-versa (this is what we mean by a "very lax" policy).

For example, if an email comes from test.sample.g0v.tw before we see an email come from sample.g0v.tw, then any email from test.sample.g0v.tw and/or sample.g0v.tw will have to wait 5 minutes from the initial arrival time of the connection. We do not make both test.sample.g0v.tw and sample.g0v.tw each wait their own 5 minute periods (our greylisting policy applies at the root domain level).

Note that greylisting does not apply to any sender on our allowlist (e.g. Meta, Amazon, Netflix, Google, Microsoft at the time of this writing).

Yes, we operate our own denylist and update it automatically in real-time and manually based off spam and malicious activity detected.

We also pull all IP addresses from the UCEPROTECT Level 1 denylist at http://wget-mirrors.uceprotect.net/rbldnsd-all/dnsbl-1.uceprotect.net.gz every hour and feed it into our denylist with a 7 day expiry.

Senders found in the denylist will receive a 421 error code (indicates to sender to retry again later) if they are not allowlisted.

By using a 421 status code instead of a 554 status code, potential false positives can be alleviated in real-time and then the message can be successfully delivered on the next attempt.

This is designed unlike other mail services, where if you are put on a blocklist, a hard and permanent failure occurs. It is often difficult to ask senders to retry messages (especially from large organizations), and therefore this approach gives roughly 5 days from the initial email attempt for either the sender, recipient, or us to step in and alleviate the issue (by requesting denylist removal).

All denylist removal requests are monitored in real-time by admins (e.g. so that recurring false positives can be permanently allowlisted by admins).

Denylist removal requests can be requested at https://forwardemail.net/denylist. Paid users have their denylist removal requests instantly processed, while non-paid users must wait for admins to process their request.

Senders that are detected to be sending spam or virus content will be added to the denylist in the following approach:

1. The initial message fingerprint is greylisted upon detection of spam or blocklist from a "trusted" sender (e.g. gmail.com, microsoft.com, apple.com).
   • If the sender was allowlisted, the message is greylisted for 1 hour.
   • If the sender is not allowlisted, the message is greylisted for 6 hours.
2. We parse denylist keys from information from the sender and message, and for each of these keys we create (if one does not already exist) a counter, increment it by 1, and cache it for 24 hours.
   • For allowlisted senders:
     • Add a key for the envelope "MAIL FROM" email address if it had passing SPF or no SPF, and it was not an admin username or no-reply username.
     • If "From" header was allowlisted, then add a key for the "From" header email address if it had passing SPF or passing and aligned DKIM.
     • If "From" header was not allowlisted, then add a key for the "From" header email address and its root parsed domain name.
   • For non-allowlisted senders:
     • Add a key for the envelope "MAIL FROM" email address if it had passing SPF.
     • If "From" header was allowlisted, then add a key for the "From" header email address if it had passing SPF or passing and aligned DKIM.
     • If "From" header was not allowlisted, then add a key for the "From" header email address and its root parsed domain name.
     • Add a key for the remote IP address of the sender.
     • Add a key for the client resolved hostname by reverse lookup from the IP address of the sender (if any).
     • Add a key for the root domain of the client resolved hostname (if any, and if it differs than the client resolved hostname).
3. If the counter reaches 5 for a non-allowlisted sender and key, then we denylist the key for 30 days and an email is sent to our abuse team. These numbers may change and updates will be reflected here as we monitor abuse.
4. If the counter reaches 10 for an allowlisted sender and key, then we denylist the key for 7 days and an email is sent to our abuse team. These numbers may change and updates will be reflected here as we monitor abuse.

NOTE: In the near future we will introduce reputation monitoring. Reputation monitoring will instead calculate when to denylist a sender based off a percentage threshold (as opposed to a rudimentary counter as noted above).

Sender rate limiting is either by the root domain parsed from a reverse PTR lookup on the sender's IP address – or if that does not yield a result, then it simply uses the sender's IP address. Note that we refer to this as Sender below.

Our MX servers have daily limits for inbound mail received for encrypted IMAP storage:

• Instead of rate limiting inbound mail received on an individual alias basis (e.g. you@yourdomain.com) – we rate limit by the alias's domain name itself (e.g. yourdomain.com). This prevents Senders from flooding the inboxes of all aliases across your domain at once.
• We have general limits that apply to all Senders across our service regardless of recipient:
  • Senders that we consider to be "trusted" as a source of truth (e.g. gmail.com, microsoft.com, apple.com) are limited to sending 100 GB per day.
  • Senders that are allowlisted are limited to sending 10 GB per day.
  • All other Senders are limited to sending 1 GB and/or 1000 messages per day.
• We have a specific limit per Sender and yourdomain.com of 1 GB and/or 1000 messages daily.

The MX servers also limit messages being forwarded to one or more recipients through rate limiting – but this only applies to Senders not on the allowlist:

• We only permit up to 100 connections per hour, per Sender resolved FQDN root domain (or) Sender remote IP address (if no reverse PTR is available), and per envelope recipient to. We store the key for rate limiting as a cryptographic hash in our Redis database.

• If you are sending email through our system, please ensure you have a reverse PTR set up for all your IP addresses (otherwise each unique FQDN root domain or IP address you send from will be rate limited).

• Note that if you send through a popular system such as Amazon SES, then you will not be rate limited since (at the time of this writing) Amazon SES is listed in our allowlist.

• If you are sending from a domain such as test.abc.123.sample.g0v.tw, then the rate limit will be imposed on sample.g0v.tw. Many spammers use hundreds of sub-domains to work around common spam filters that only rate limit unique hostnames as opposed to unique FQDN root domains.

• Senders that exceed the rate limit will be rejected with a 421 error.

Our IMAP and SMTP servers limit your aliases from having more than 60 concurrent connections at once.

Our MX servers limit non-allowlisted senders from establishing more than 10 concurrent connections (with 3 minute cache expiry for the counter, which mirrors our socket timeout of 3 minutes).

Misdirected bounces or bounce spam (known as "Backscatter") can cause negative reputation to sender IP addreses.

We take two steps to protect against backscatter, which is detailed in the following sections Prevent bounces from known MAIL FROM spammers and Prevent unnecessary bounces to protect against backscatter below.

Prevent bounces from known MAIL FROM spammers

We pull the list from Backscatter.org (powered by UCEPROTECT) at http://wget-mirrors.uceprotect.net/rbldnsd-all/ips.backscatterer.org.gz every hour and feed it into our Redis database (we also compare the difference in advance; in case any IP's were removed that need to be honored).

If the MAIL FROM is blank OR is equal to (case-insensitive) any of the following usernames (the portion before the @ in an email), then we check to see if the sender IP matches one from this list.

• automailer
• autoresponder
• bounce
• bounce-notification
• bounce-notifications
• bounces
• e-bounce
• ebounce
• host-master
• host.master
• hostmaster
• localhost
• mail-daemon
• mail.daemon
• maildaemon
• mailer
• mailer-daemon
• mailer.daemon
• mailerdaemon
• post-master
• post.master
• postmaster

If the sender's IP is listed (and not in our allowlist), then we send a 554 error with the message The IP ${session.remoteAddress} is blocked by https://www.backscatterer.org/index.php?target=test&ip=${session.remoteAddress}. We will be alerted if a sender is on both the Backscatterer list and in our allowlist so we can resolve the issue if necessary.

The techniques described in this section adhere to the "SAFE MODE" recommendation at https://www.backscatterer.org/?target=usage – where we only check the sender IP if certain conditions have already been met.

Prevent unnecessary bounces to protect against backscatter

Bounces are emails that indicate email forwarding completely failed to the recipient and the email will not be retried.

A common reason for getting listed on the Backscatterer list is misdirected bounces or bounce spam, so we must protect against this in a few ways:

1. We only send bounces when >= 500 status code errors occur (when emails attempted to be forwarded have failed, e.g. Gmail responds with a 500 level error).

2. We only send bounces once and once only (we use a calculated bounce fingerprint key and store it in cache to prevent sending duplicates). The bounce fingerprint is a key that is the message's fingerprint combined with a hash of the bounce address and its error code). See the section on Fingerprinting for more insight into how the message fingerprint is calculated. Successfully sent bounce fingerprints will expire after 7 days in our Redis cache.

3. We only send bounces when the MAIL FROM is not blank and does not contain (case-insensitive) one of the following usernames (the portion before the @ in an email). Note that this list is a little bit shorter than the one above in the MAIL FROM check because we don't want to have false positives (e.g. security@ is a valid address that you might want to get a bounce for; a lot of folks use security@ for their bug bounty programs).

   • abuse@
   • mailer-daemon@
   • mailer_daemon@
   • mailerdaemon@

4. We don't send bounces if the original message had any of the following headers (case-insensitive):

   • Auto-Submitted (with a value of no)
   • X-Auto-Response-Suppress (with a value of dr, autoreply, auto-reply, auto_reply, or all)
   • List-Id
   • List-Unsubscribe
   • Feedback-ID
   • X-Auto-Reply
   • X-Autoreply
   • X-Auto-Respond
   • X-Autorespond
   • Precedence (with a value of bulk, autoreply, auto-reply, auto_reply, or list)

An email's fingerprint is used for determining uniqueness of an email and to prevent duplicate messages from being delivered and duplicate bounces from being sent.

The fingerprint is calculated from the following list:

• Client resolved FQDN hostname or IP address
• Message-ID header value (if any)
• Date header value (if any)
• From header value (if any)
• To header value (if any)
• Cc header value (if any)
• Subject header value (if any)
• Body value (if any)

If you see this error message in Gmail when you send a test to yourself, or when a person you're emailing with your alias sees an email from you for the first time, then please do not worry – as this is a built-in safety feature of Gmail.

You can simply click "Looks safe". For example, if you were to send a test message using the send mail as feature (to someone else), then they will not see this message.

However if they do see this message, it's because they were normally used to seeing your emails come from john@gmail.com instead of john@customdomain.com (just an example). Gmail will alert the users just to make sure things are safe just in case, there is no workaround.

This topic is related to a widely known issue in Gmail where extra info appears next to a sender's name.

As of May 2023 we support sending email with SMTP as an add-on for all paid users – which means that you can remove the via forwardemail dot net in Gmail.

Note that this FAQ topic is specific for those using the How to Send Mail As using Gmail feature.

Please see the section on Do you support sending email with SMTP for configuration instructions.

Yes, as of May 5, 2020 we have added this feature. Right now the feature is domain-specific, as opposed to alias-specific. If you require it to be alias-specific, please contact us to let us know of your needs.

If you are on the free plan, then simply add a new DNS TXT record as shown below, but change the port from 25 to the port of your choosing.

For example, if I want all emails that go to sample.g0v.tw to forward to alias recipients' SMTP port of 1337 instead of 25:

Yes! Automatic refunds occur when you upgrade, downgrade, or cancel your account within 30-days from when your plan first started. This only applies for first-time customers.

We do not pro-rate nor refund the difference when you switch plans. Instead we convert the remaining duration from your existing plan's expiration date into the closest relative duration for your new plan (rounded down by month).

Note that if you upgrade or downgrade between paid plans within a 30-day window since first starting a paid plan, then we will automatically refund the full amount from your existing plan.

Yes, as of August 14, 2024 we have added this feature. You can now go to My Account → Domains → Settings → Bounce Webhook URL and configure an http:// or https:// URL that we will send a POST request to whenever outbound SMTP emails bounce.

This is useful for you to manage and monitor your outbound SMTP – and can be used to maintain subscribers, opt-out, and detect whenever bounces occur.

Bounce webhook payloads are sent as a JSON with these properties:

• email_id (String) - email ID that corresponds to an email in My Account → Emails (outbound SMTP)
• list_id (String) - the List-ID header (case-insensitive) value, if any, from the original outbound email
• list_unsubscribe (String) - the List-Unsubscribe header (case-insensitive) value, if any, from the original outbound email
• feedback_id (String) - the Feedback-ID header (case-insensitive) value, if any, from the original outbound email
• recipient (String) - the email address of the recipient that bounced or errored
• message (String) - a detailed error message for the bounce
• response (String) - the SMTP response message
• response_code (Number) - the parsed SMTP response code
• truth_source (String) - if the response code was from a trusted source, this value will be populated with the root domain name (e.g. google.com or yahoo.com)
• bounce (Object) - an object containing the following properties that detail the bounce and rejection status
  • action (String) - bounce action (e.g. "reject")
  • message (String) - bounce reason (e.g. "Message Sender Blocked By Receiving Server")
  • category (String) - bounce category (e.g. "block")
  • code (Number) - bounce status code (e.g. 554)
  • status (String) - bounce code from response message (e.g. 5.7.1)
  • line (Number) - parsed line number, if any, from Zone-MTA bounce parse list (e.g. 526)
• headers (Object) - key value pair of headers for the outbound email
• bounced_at (String) - ISO 8601 formatted Date for when the bounce error occurred

For example:

{
  "email_id": "66bcce793ef7b2a0928e14ba",
  "recipient": "example@gmail.com",
  "message": "The email account that you tried to reach is over quota.",
  "response": "552 5.2.2 The email account that you tried to reach is over quota.",
  "response_code": 552,
  "truth_source": "google.com",
  "bounce": {
    "action": "reject",
    "message": "Gmail Mailbox is full",
    "category": "capacity",
    "code": 552,
    "status": "5.2.2",
    "line": 300
  },
  "headers": {},
  "bounced_at": "2024-08-24T01:50:02.828Z"
}

Here are a few additional notes regarding bounce webhooks:

• If the webhook payload contains a list_id, list_unsubscribe, or feedback_id value, then you should take appropriate action to remove the recipient from the list if necessary.
  • If the bounce.category value was one "block", "recipient", "spam", or "virus", then you should definitely remove the user from the list.
• If you need to verify webhook payloads (to ensure they're actually coming from our server), then you can resolve the remote client IP address client hostname using a reverse lookup – it should be smtp.forwardemail.net.
  • You can also check the IP against our published IP addresses.
  • Go to My Account → Domains → Settings → Webhook Signature Payload Verification Key to obtain your webhook key.
    • You can rotate this key at anytime for security reasons.
    • Calculate and compare the X-Webhook-Signature value from our webhook request with the computed body value using this key. An example of how to do this is available at this Stack Overflow post.
  • See the discussion at https://github.com/forwardemail/free-email-forwarding/issues/235 for more insight.
• We will wait for up to 5 seconds for your webhook endpoint to respond with a 200 status code, and we will retry up to 1 time.
• If we detect that your bounce webhook URL has an error while we try to send a request to it, then we will send you a courtesy email once a week.

Yes, as of May 15, 2020 we have added this feature. You can simply add webhook(s) exactly like you would with any recipient! Please ensure that you have the "http" or "https" protocol prefixed in the webhook's URL.

If you are on the free plan, then simply add a new DNS TXT record as shown below:

For example, if I want all emails that go to alias@sample.g0v.tw to forward to a new request bin test endpoint:

Or perhaps you want all emails that go to sample.g0v.tw to forward to this endpoint:

Here are additional notes regarding webhooks:

• If you need to verify webhook payloads (to ensure they're actually coming from our server), then you can resolve the remote client IP address client hostname using a reverse lookup – it should be either mx1.forwardemail.net or mx2.forwardemail.net.

  • You can also check the IP against our published IP addresses.
  • If you're on a paid plan, then go to My Account → Domains → Settings → Webhook Signature Payload Verification Key to obtain your webhook key.
    • You can rotate this key at anytime for security reasons.
    • Calculate and compare the X-Webhook-Signature value from our webhook request with the computed body value using this key. An example of how to do this is available at this Stack Overflow post.
  • See the discussion at https://github.com/forwardemail/free-email-forwarding/issues/235 for more insight.

• If a webhook does not respond with a 200 status code, then we will store its response in the error log created – which is useful for debugging.

• Webhook HTTP requests will retry up to 3 times every SMTP connection attempt, with a 60 second max timeout per endpoint POST request. Note that this does not mean that it only retries 3 times, it will actually retry continously over time by sending a SMTP code of 421 (which indicates to the sender retry later) after the 3rd failed HTTP POST request attempt. This means the email will retry continuously for days until a 200 status code is achieved.

• We will retry automatically based off the default status and error codes used in superagent's retry method (we are maintainers).

• We group together webhook HTTP requests to the same endpoint in one request instead of multiple) in order to save resources and speed up response time. For example, if you send an email to webhook1@sample.g0v.tw, webhook2@sample.g0v.tw, and webhook3@sample.g0v.tw, and all of these are configured to hit the same exact endpoint URL, then only one request will be made. We group together by exact endpoint matching with strict equality.

• Note that we use the mailparser library's "simpleParser" method to parse the message into a JSON friendly object.

• Raw email value as a String is given as the property "raw".

• Authentication results are given as properties "dkim", "spf", "arc", "dmarc", and "bimi".

• The parsed email headers is given as the property "headers" – but also note you can use "headerLines" for easier iteration and parsing.

• The grouped recipients for this webhook are grouped together and given as the property "recipients".

• The SMTP session information is given as the property "session". This contains information about the sender of the message, arrival time of the message, HELO, and client hostname. The client hostname value as session.clientHostname is either the FQDN (from a reverse PTR lookup) or it is session.remoteAddress wrapped in brackets (e.g. "[127.0.0.1]").

• If you need a quick way to get the value of X-Original-To, then you can use the value of session.recipient (see example below). The header X-Original-To is a header we add to messages for debugging with the original RCPT TO email address for the message.

• If there are attachments, they will be appended to the attachments Array with Buffer values. You can parse them back into content using an approach with JavaScript such as:

  const data = [
    104,
    101,
    108,
    108,
    111,
    32,
    119,
    111,
    114,
    108,
    100,
    33
  ];
  
  //
  // outputs "hello world!" to the console
  // (this is the content from the filename "text1.txt" in the example JSON request payload above)
  //
  console.log(Buffer.from(data).toString());
  

{
  "attachments": [
    {
      "type": "attachment",
      "content": {
        "type": "Buffer",
        "data": [
          104,
          101,
          108,
          108,
          111,
          32,
          119,
          111,
          114,
          108,
          100,
          33
        ]
      },
      "contentType": "text/plain",
      "partId": "2",
      "release": null,
      "contentDisposition": "attachment",
      "filename": "text1.txt",
      "headers": {},
      "checksum": "fc3ff98e8c6a0d3087d515c0473f8677",
      "size": 12
    }
  ],
  "headers": "ARC-Seal: i=1; a=rsa-sha256; t=1653506802; cv=none; d=forwardemail.net;\r\n s=default;\r\n b=R6QJ0tGwwjg2VPxiAlVIKxsg3jEPtRGKPTIOdZNWuhWrbssttFdOYzRRqvacDyN5SLoyDhVye\r\n DUA/64IxANXdHVFlpR258Yp7WxLDv2gtJD5vNSKYmUJZOWk1TynmlqTYrp0Vuqg2xIUjIlPBWAJ\r\n PPNx4JvOLjJuWYynU2qIWz0=\r\nARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed;\r\n d=forwardemail.net; h=MIME-Version: Date: Message-ID: From: Content-Type;\r\n q=dns/txt; s=default; t=1653506802;\r\n bh=cEYDoyTy+Ub29XZt/zXR+sprfUE6BW0y5cHfah01PT4=;\r\n b=F/t56AAXr2Kv3G6VsbdT5OKDVJf2ulhwLiTM18Ra4tDPUKPSGSLKrWvxiXEg5NMWwdWnsOYrL\r\n r3YSm4uMxVMhHZbHm/sUu4QZq5/18hQsAkCv6fI9ifTjDwBrN5zpLOhPoZFFo+TyvHxiII3Xv3L\r\n UEzmUIIaJRX6tboQ160tino=\r\nARC-Authentication-Results: i=1; mx1.forwardemail.net;\r\n dkim=none (message not signed);\r\n spf=none (mx1.forwardemail.net: example.net does not designate permitted sender hosts) smtp.mailfrom=test@example.net smtp.helo=user.oem.local;\r\n dmarc=none header.from=sample.g0v.tw;\r\n bimi=skipped (DMARC not enabled)\r\nReceived-SPF: none (mx1.forwardemail.net: example.net does not designate permitted sender hosts) client-ip=127.0.0.1;\r\nAuthentication-Results: mx1.forwardemail.net;\r\n dkim=none (message not signed);\r\n spf=none (mx1.forwardemail.net: example.net does not designate permitted sender hosts) smtp.mailfrom=test@example.net smtp.helo=user.oem.local;\r\n dmarc=none header.from=sample.g0v.tw;\r\n bimi=skipped (DMARC not enabled)\r\n",
  "headerLines": [
    {
      "key": "arc-seal",
      "line": "ARC-Seal: i=1; a=rsa-sha256; t=1653506802; cv=none; d=forwardemail.net;\r\n s=default;\r\n b=R6QJ0tGwwjg2VPxiAlVIKxsg3jEPtRGKPTIOdZNWuhWrbssttFdOYzRRqvacDyN5SLoyDhVye\r\n DUA/64IxANXdHVFlpR258Yp7WxLDv2gtJD5vNSKYmUJZOWk1TynmlqTYrp0Vuqg2xIUjIlPBWAJ\r\n PPNx4JvOLjJuWYynU2qIWz0="
    },
    {
      "key": "arc-message-signature",
      "line": "ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed;\r\n d=forwardemail.net; h=MIME-Version: Date: Message-ID: From: Content-Type;\r\n q=dns/txt; s=default; t=1653506802;\r\n bh=cEYDoyTy+Ub29XZt/zXR+sprfUE6BW0y5cHfah01PT4=;\r\n b=F/t56AAXr2Kv3G6VsbdT5OKDVJf2ulhwLiTM18Ra4tDPUKPSGSLKrWvxiXEg5NMWwdWnsOYrL\r\n r3YSm4uMxVMhHZbHm/sUu4QZq5/18hQsAkCv6fI9ifTjDwBrN5zpLOhPoZFFo+TyvHxiII3Xv3L\r\n UEzmUIIaJRX6tboQ160tino="
    },
    {
      "key": "arc-authentication-results",
      "line": "ARC-Authentication-Results: i=1; mx1.forwardemail.net;\r\n dkim=none (message not signed);\r\n spf=none (mx1.forwardemail.net: example.net does not designate permitted sender hosts) smtp.mailfrom=test@example.net smtp.helo=user.oem.local;\r\n dmarc=none header.from=sample.g0v.tw;\r\n bimi=skipped (DMARC not enabled)"
    },
    {
      "key": "received-spf",
      "line": "Received-SPF: none (mx1.forwardemail.net: example.net does not designate permitted sender hosts) client-ip=127.0.0.1;"
    },
    {
      "key": "authentication-results",
      "line": "Authentication-Results: mx1.forwardemail.net;\r\n dkim=none (message not signed);\r\n spf=none (mx1.forwardemail.net: example.net does not designate permitted sender hosts) smtp.mailfrom=test@example.net smtp.helo=user.oem.local;\r\n dmarc=none header.from=sample.g0v.tw;\r\n bimi=skipped (DMARC not enabled)"
    },
    {
      "key": "x-forwardemail-sender",
      "line": "X-ForwardEmail-Sender: rfc822; test@example.net"
    },
    {
      "key": "x-forwardemail-session-id",
      "line": "X-ForwardEmail-Session-ID: w2czxgznghn5ryyw"
    },
    {
      "key": "x-forwardemail-version",
      "line": "X-ForwardEmail-Version: 9.0.0"
    },
    {
      "key": "content-type",
      "line": "Content-Type: multipart/mixed; boundary=\"--_NmP-179a735428ca7575-Part_1\""
    },
    {
      "key": "from",
      "line": "From: some <random@sample.g0v.tw>"
    },
    {
      "key": "message-id",
      "line": "Message-ID: <69ad5fc2-91cb-728f-ae5c-eeedc5f267b6@example.net>"
    },
    {
      "key": "date",
      "line": "Date: Wed, 25 May 2022 19:26:41 +0000"
    },
    {
      "key": "mime-version",
      "line": "MIME-Version: 1.0"
    }
  ],
  "html": "<strong>some random text</strong>",
  "text": "some random text",
  "textAsHtml": "<p>some random text</p>",
  "date": "2022-05-25T19:26:41.000Z",
  "from": {
    "value": [
      {
        "address": "random@sample.g0v.tw",
        "name": "some"
      }
    ],
    "html": "<span class=\"mp_address_group\"><span class=\"mp_address_name\">some</span> &lt;<a href=\"mailto:random@sample.g0v.tw\" class=\"mp_address_email\">random@sample.g0v.tw</a>&gt;</span>",
    "text": "some <random@sample.g0v.tw>"
  },
  "messageId": "<69ad5fc2-91cb-728f-ae5c-eeedc5f267b6@example.net>",
  "raw": "ARC-Seal: i=1; a=rsa-sha256; t=1653506802; cv=none; d=forwardemail.net;\r\n s=default;\r\n b=R6QJ0tGwwjg2VPxiAlVIKxsg3jEPtRGKPTIOdZNWuhWrbssttFdOYzRRqvacDyN5SLoyDhVye\r\n DUA/64IxANXdHVFlpR258Yp7WxLDv2gtJD5vNSKYmUJZOWk1TynmlqTYrp0Vuqg2xIUjIlPBWAJ\r\n PPNx4JvOLjJuWYynU2qIWz0=\r\nARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed;\r\n d=forwardemail.net; h=MIME-Version: Date: Message-ID: From: Content-Type;\r\n q=dns/txt; s=default; t=1653506802;\r\n bh=cEYDoyTy+Ub29XZt/zXR+sprfUE6BW0y5cHfah01PT4=;\r\n b=F/t56AAXr2Kv3G6VsbdT5OKDVJf2ulhwLiTM18Ra4tDPUKPSGSLKrWvxiXEg5NMWwdWnsOYrL\r\n r3YSm4uMxVMhHZbHm/sUu4QZq5/18hQsAkCv6fI9ifTjDwBrN5zpLOhPoZFFo+TyvHxiII3Xv3L\r\n UEzmUIIaJRX6tboQ160tino=\r\nARC-Authentication-Results: i=1; mx1.forwardemail.net;\r\n dkim=none (message not signed);\r\n spf=none (mx1.forwardemail.net: example.net does not designate permitted sender hosts) smtp.mailfrom=test@example.net smtp.helo=user.oem.local;\r\n dmarc=none header.from=sample.g0v.tw;\r\n bimi=skipped (DMARC not enabled)\r\nReceived-SPF: none (mx1.forwardemail.net: example.net does not designate permitted sender hosts) client-ip=127.0.0.1;\r\nAuthentication-Results: mx1.forwardemail.net;\r\n dkim=none (message not signed);\r\n spf=none (mx1.forwardemail.net: example.net does not designate permitted sender hosts) smtp.mailfrom=test@example.net smtp.helo=user.oem.local;\r\n dmarc=none header.from=sample.g0v.tw;\r\n bimi=skipped (DMARC not enabled)\r\nX-ForwardEmail-Sender: rfc822; test@example.net\r\nX-ForwardEmail-Session-ID: w2czxgznghn5ryyw\r\nX-ForwardEmail-Version: 9.0.0\r\nContent-Type: multipart/mixed; boundary=\"--_NmP-179a735428ca7575-Part_1\"\r\nFrom: some <random@sample.g0v.tw>\r\nMessage-ID: <69ad5fc2-91cb-728f-ae5c-eeedc5f267b6@example.net>\r\nDate: Wed, 25 May 2022 19:26:41 +0000\r\nMIME-Version: 1.0\r\n\r\n----_NmP-179a735428ca7575-Part_1\r\nContent-Type: multipart/alternative;\r\n boundary=\"--_NmP-179a735428ca7575-Part_2\"\r\n\r\n----_NmP-179a735428ca7575-Part_2\r\nContent-Type: text/plain; charset=utf-8\r\nContent-Transfer-Encoding: 7bit\r\n\r\nsome random text\r\n----_NmP-179a735428ca7575-Part_2\r\nContent-Type: text/html; charset=utf-8\r\nContent-Transfer-Encoding: 7bit\r\n\r\n<strong>some random text</strong>\r\n----_NmP-179a735428ca7575-Part_2--\r\n\r\n----_NmP-179a735428ca7575-Part_1\r\nContent-Type: text/plain; name=text1.txt\r\nContent-Transfer-Encoding: base64\r\nContent-Disposition: attachment; filename=text1.txt\r\n\r\naGVsbG8gd29ybGQh\r\n----_NmP-179a735428ca7575-Part_1--\r\n",
  "dkim": {
    "headerFrom": [
      "random@sample.g0v.tw"
    ],
    "envelopeFrom": "test@example.net",
    "results": [
      {
        "status": {
          "result": "none",
          "comment": "message not signed"
        },
        "info": "dkim=none (message not signed)"
      }
    ]
  },
  "spf": {
    "domain": "example.net",
    "client-ip": "127.0.0.1",
    "helo": "user.oem.local",
    "envelope-from": "test@example.net",
    "status": {
      "result": "none",
      "comment": "mx1.forwardemail.net: example.net does not designate permitted sender hosts",
      "smtp": {
        "mailfrom": "test@example.net",
        "helo": "user.oem.local"
      }
    },
    "header": "Received-SPF: none (mx1.forwardemail.net: example.net does not designate permitted sender hosts) client-ip=127.0.0.1;",
    "info": "spf=none (mx1.forwardemail.net: example.net does not designate permitted sender hosts) smtp.mailfrom=test@example.net smtp.helo=user.oem.local",
    "lookups": {
      "limit": 50,
      "count": 1
    }
  },
  "arc": {
    "status": {
      "result": "none"
    },
    "i": 0,
    "authResults": "mx1.forwardemail.net;\r\n dkim=none (message not signed);\r\n spf=none (mx1.forwardemail.net: example.net does not designate permitted sender hosts) smtp.mailfrom=test@example.net smtp.helo=user.oem.local;\r\n dmarc=none header.from=sample.g0v.tw;\r\n bimi=skipped (DMARC not enabled)"
  },
  "dmarc": {
    "status": {
      "result": "none",
      "header": {
        "from": "sample.g0v.tw"
      }
    },
    "domain": "sample.g0v.tw",
    "info": "dmarc=none header.from=sample.g0v.tw"
  },
  "bimi": {
    "status": {
      "header": {},
      "result": "skipped",
      "comment": "DMARC not enabled"
    },
    "info": "bimi=skipped (DMARC not enabled)"
  },
  "recipients": [
    "webhook1@webhooks.net"
  ],
  "session": {
    "recipient": "webhook1@webhooks.net",
    "remoteAddress": "127.0.0.1",
    "remotePort": 65138,
    "clientHostname": "[127.0.0.1]",
    "hostNameAppearsAs": "user.oem.local",
    "sender": "test@example.net",
    "mta": "mx1.forwardemail.net",
    "arrivalDate": "2022-05-25T19:26:41.423Z",
    "arrivalTime": 1653506801423
  }
}

Yes, as of September 27, 2021 we have added this feature. You can simply write regular expressions ("regex") for matching aliases and performing substitions.

Regular expression supported aliases are ones that start with a / and end with / and their recipients are email addresses or webhooks. The recipients can also include regex substitution support (e.g. $1, $2).

We support two regular expression flags including i and g. The case-insensitive flag of i is a permanent default and it is always enforced. The global flag of g can be added by you by affixing the ending / with /g.

Note that we also support our disabled alias feature for the recipient portion with our regex support.

Regular expressions are not supported on global vanity domains (as this could be a security vulnerability).

If you are on the free plan, then simply add a new DNS TXT record using one or more of the provided examples below:

No, it is not recommended, as you can only use one mail exchange server at a time. Fallbacks are usually never retried due to priority misconfigurations and mail servers not respecting MX exchange priority checking.

Yes, simply edit your DNS TXT record and prefix the alias with either one, two, or three exclamation marks (see below).

Note that you should preserve the ":" mapping, as this is required if you ever decide to toggle this off (and it's also used for importing if you upgrade to one of our paid plans).

For quiet reject (appears to sender as if the message was sent successfully, but actually goes nowhere) (status code 250): If you prefix an alias with "!" (single exclamation mark) then it will return a successful status code of 250 to senders attempting to send to this address, but the emails themselves will go nowhere (e.g. a blackhole or /dev/null).

For soft reject (status code 421): If you prefix an alias with "!!" (double exclamation mark) then it will return a soft error status code of 421 to senders attempting to send to this address, and the emails will often be retried for up to 5 days before rejection and bounce.

For hard reject (status code 550): If you prefix an alias with "!!!" (triple exclamation mark) then it will return a permanent error status code of 550 to senders attempting to send to this address and the emails will be rejected and bounce.

For example, if I want all emails that go to alias@sample.g0v.tw to stop flowing through to user@gmail.com and get rejected and bounce (e.g. use three exclamation marks):

Yes, absolutely. Just specify multiple recipients in your TXT records.

For example, if I want an email that goes to hello@sample.g0v.tw to get forwarded to user+a@gmail.com and user+b@gmail.com, then my TXT record would look like this:

Or, you could specify them in two separate lines, such as this:

It's up to you!

Yes, you can. Just specify multiple global catch-all recipients in your TXT records.

For example, if I want every email that goes to *@sample.g0v.tw (the asterisk meaning its a wildcard aka catch-all) to get forwarded to user+a@gmail.com and user+b@gmail.com, then my TXT record would look like this:

Or, you could specify them in two separate lines, such as this:

It's up to you!

Yes, the default limit is 10. This does NOT mean that you can only have 10 aliases on your domain name. You can have as many aliases as you want (an unlimited amount). It means that you can only forward one alias to 10 unique email addresses. You could have hello:user+1@gmail.com, hello:user+2@gmail.com, hello:user+3@gmail.com, … (from 1-10) – and any emails to hello@sample.g0v.tw would get forwarded to user+1@gmail.com, user+2@gmail.com, user+3@gmail.com, … (from 1-10).

Yes, you can, however you still must adhere to the maximum limit. If you have hello:linus@sample.g0v.tw and linus:user@gmail.com, then emails to hello@sample.g0v.tw would get forwarded to linus@sample.g0v.tw and user@gmail.com. Note that an error will be thrown if you attempt to recursively forward emails beyond the maximum limit.

We use MX and TXT record verification, therefore if you add this service's respective MX and TXT records, then you're registered. If you remove them, then you're unregistered. You have ownership of your domain and DNS management, so if someone has access to that then that's a problem.

The service continues to run thanks to users that upgraded to paid plans. We want to provide a free alternative for people that are using closed-source forwarding services (and subsequently risking their privacy and security).

We default to a 50MB size limit, which includes content, headers, and attachments. Note that services such as Gmail and Outlook allow only 25MB size limit, and if you exceed the limit when sending to addresses at those providers you will receive an error message.

An error with the proper response code is returned if the file size limit is exceeded.

No, we do not write to disk or store logs – with the exception of errors and outbound SMTP (see our Privacy Policy).

Everything is done in-memory and our source code is on GitHub.

No, we do not write to disk or store logs – with the exception of errors and outbound SMTP (see our Privacy Policy).

Everything is done in-memory and our source code is on GitHub.

Yes. You can access error logs under My Account → Logs or My Account → Domains.

As of February 2023, we store error logs for 4xx and 5xx SMTP response codes for a period of 7 days – which contain the SMTP error, envelope, and email headers (we do not store the email body nor attachments).

Error logs allow you to check for missing important emails and mitigate spam false positives for your domains. They are also a great resource for debugging issues with email webhooks (since the error logs contain the webhook endpoint response).

Error logs for rate limiting and greylisting are not accessible since the connection ends early (e.g. before RCPT TO and MAIL FROM commands can be transmitted).

See our Privacy Policy for more insight.

No, absolutely not. See our Privacy Policy.

Many other email forwarding services store and could potentially read your email. There is no reason why forwarded emails need to be stored to disk storage – and therefore we architected the first open-source solution that does it all in-memory.

We believe you should have a right to privacy and we strictly respect it. The code that is deployed to the server is open-source software on GitHub for transparency and to build trust.

Yes, absolutely.

Yes, absolutely. Instead of using "@", ".", or blank as the name/host/alias, you just use the sub-domain name as the value instead.

If you want foo.sample.g0v.tw to forward emails, then enter foo as the name/host/alias value in your DNS settings (for both MX and TXT records).

Yes, absolutely.

Yes! As of December 13, 2023 we have added support for passkeys due to high demand.

Passkeys allow you to securely log in without requiring a password and two-factor authentication.

You can validate your identity with touch, facial recognition, device-based password, or PIN.

We allow you to manage up to 30 passkeys at once, so that you can log in with all of your devices with ease.

Learn more about passkeys at the following links:

• Sign-in to your applications and websites with passkeys (Google)
• Use passkeys to sign in to apps and websites on iPhone (Apple)
• Wikipedia article on Passkeys

Yes, it has tests written with ava and also has code coverage.

Yes, absolutely. For example if you're sending an email to hello@sample.g0v.tw and it's registered to forward to user@gmail.com, then the SMTP response message and code from the "gmail.com" SMTP server will be returned instead of the proxy server at "mx1.forwardemail.net" or "mx2.forwardemail.net".

See our sections on How does your email forwarding system work, How do you handle email delivery issues, and How do you handle your IP addresses becoming blocked above.

This section guides you if your outbound mail is using our SMTP servers (e.g. smtp.forwardemail.net) (or forwarded via mx1.forwardemail.net or mx2.forwardemail.net) and it is being delivered in the Spam or Junk folder of recipients.

We routinely monitor our IP addresses against all reputable DNS denylists, therefore it is most likely a domain-reputation specific issue.

You can try to use one or more of these tools to check your domain's reputation and categorization:

• Cloudflare Domain Categorization Feedback
• Spamhaus IP and Domain Reputation Checker
• Cisco Talos IP and Domain Reputation Center
• Barracuda IP and Domain Reputation Lookup
• MX Toolbox Blacklist Check
• Google Postmaster Tools
• Yahoo Sender Hub (includes Verizon/AOL)
• MultiRBL.valli.org Blacklist Check

If you need additional help or find that we are false-positive listed as spam by a certain email service provider, then please contact us.

You should unsubscribe from the emailing list (if possible) and block the sender.

Please do not report the message as spam, but instead forward it to our manually curated and privacy-focused abuse prevention system.

The email address to forward spam to is: abuse@forwardemail.net

Yes! As of October 2, 2018 we have added this feature. See How to Send Mail As using Gmail above!

You should also set the SPF record for Gmail in your DNS configuration TXT record.

Yes! As of October 2, 2018 we have added this feature. Simply view these two links from Microsoft below:

• https://support.office.com/en-us/article/add-or-remove-an-email-alias-in-outlook-com-459b1989-356d-40fa-a689-8f285b13f1f2
• https://support.office.com/en-us/article/send-email-from-a-different-address-in-outlook-com-ccba89cb-141c-4a36-8c56-6d16a8556d2e

You should also set the SPF record for Outlook in your DNS configuration TXT record.

Unfortunately Apple does not allow this, regardless of which service you use. However you can use the Mail app along with your domain's email account.

• https://discussions.apple.com/thread/8316291
• https://discussions.apple.com/thread/6876839

Yes, however "relatively unknown" senders are rate limited to 100 connections per hour per hostname or IP. See the section on Rate Limiting and Greylisting above.

By "relatively unknown", we mean senders that do not appear in the allowlist.

If this limit is exceeded we send a 421 response code which tells the senders mail server to retry again later.

If you're using Gmail, then follow these steps below:

1. Go to https://google.com and sign out of all email accounts
2. Click "Sign In" and on the drop-down click on "other account"
3. Select "Use another account"
4. Select "Create account"
5. Select "Use my current email address instead"
6. Enter your custom domain name email address
7. Retrieve the verification email sent to your email address
8. Enter the verification code from this email
9. Complete profile information for your new Google account
10. Agree to all Privacy and Terms of Use policies
11. Go to https://google.com and in the top right corner, click on your profile icon, and click on the "change" button
12. Upload a new photo or avatar for your account
13. Changes will take approximately 1-2 hours to propagate, but sometimes may be very quick.
14. Send a test email and the profile photo should appear.

Yes, as of March 2, 2023 we support MTA-STS. You can use this template if you wish to enable it on your domain.

Our configuration can be found publicly on GitHub at https://github.com/forwardemail/mta-sts.forwardemail.net.

Yes. We have built-in support for SPF, DKIM, DMARC, ARC, and SRS across all plans. We have also worked extensively with the original authors of these specifications and other email experts to ensure perfection and high deliverability.

Yes. Regardless of which plan you are on, you will pay only one monthly rate – which covers all of your domains.

We accept cards, wallets, and bank transfers using Stripe and PayPal – for one-time payments or monthly, quarterly, or yearly subscriptions.

We created an open-source software project 🍊 Tangerine and use it for DNS lookups. The default DNS servers used are 1.1.1.1 and 1.0.0.1, and DNS queries are through DNS over HTTPS ("DoH") at the application layer.

🍊 Tangerine uses CloudFlare's privacy-first consumer DNS service by default.