# Download page should include SHA256 hash values

**URL:** https://forum.torproject.org/t/download-page-should-include-sha256-hash-values/21230
**Category:** Websites Feedback
**Created:** [February 13, 2026, 10:11am UTC](https://forum.torproject.org/t/download-page-should-include-sha256-hash-values/21230 "2026-02-13T10:11:58Z")
**Posts on this page:** 11
**Page:** 1

<div class="post-metadata">

### Author: ![compis](https://forum.torproject.org/letter_avatar_proxy/v4/letter/c/9d8465/32.png) [@compis](https://forum.torproject.org/u/compis)
#### Post date: [February 13, 2026, 10:11am UTC](https://forum.torproject.org/t/download-page-should-include-sha256-hash-values/21230/1 "2026-02-13T10:11:58Z")

</div>

The Tor download page, [Tor Project | Download](https://www.torproject.org/download/) should include the SHA256 hash values next to signatures.

---

<div class="post-metadata">

### Author: ![Lind](https://forum.torproject.org/user_avatar/forum.torproject.org/lind/32/7525_2.png) [@Lind](https://forum.torproject.org/u/Lind)
#### Post date: [February 15, 2026, 12:49pm UTC](https://forum.torproject.org/t/download-page-should-include-sha256-hash-values/21230/2 "2026-02-15T12:49:02Z")

</div>

I don’t get the appeal of adding hash values additional to signatures. For advanced users that wants an additional security guarantee, using signature is better than a string of hash value, because they’re cryptographically linked to the Tor project team. For casual users, they just don’t verify and use whatever they downloaded.

---

<div class="post-metadata">

### Author: ![BobbyB](https://forum.torproject.org/user_avatar/forum.torproject.org/bobbyb/32/5456_2.png) [@BobbyB](https://forum.torproject.org/u/BobbyB)
#### Post date: [February 16, 2026, 10:01am UTC](https://forum.torproject.org/t/download-page-should-include-sha256-hash-values/21230/3 "2026-02-16T10:01:57Z")

</div>

I disagree. Not everyone has the utilities (or wants the utilities) to check the sig using the asc file like I do. But I do agree that most will just use whatever they download.

I’m assuming the original writer is using Windows so right click on the file name and check the digital signature using properties. Windows uses it own trust chains to verify.

---

<div class="post-metadata">

### Author: ![Preetor681](https://forum.torproject.org/letter_avatar_proxy/v4/letter/p/a587f6/32.png) [@Preetor681](https://forum.torproject.org/u/Preetor681)
#### Post date: [February 16, 2026, 10:02am UTC](https://forum.torproject.org/t/download-page-should-include-sha256-hash-values/21230/4 "2026-02-16T10:02:32Z")

</div>

This is purely my personal opinion, but the value of displaying hashes on the page is greatly overrated.  
Signatures, yes, make sense, provided the key is downloaded through a secure channel.  
But hashes present a completely unclear threat model. An adversary can hack TLS and replace the downloaded file, but they can’t change the hash displayed on the page. Where did it come from?  
An adversary has root access to TorProject servers and can replace files with malicious ones, but can’t change the displayed hash on the page?

---

<div class="post-metadata">

### Author: ![BobbyB](https://forum.torproject.org/user_avatar/forum.torproject.org/bobbyb/32/5456_2.png) [@BobbyB](https://forum.torproject.org/u/BobbyB)
#### Post date: [February 17, 2026, 12:42am UTC](https://forum.torproject.org/t/download-page-should-include-sha256-hash-values/21230/5 "2026-02-17T00:42:57Z")

</div>

OK I don’t get this.

> [@Preetor681](#):
>
> But hashes present a completely unclear threat model. An adversary can hack TLS and replace the downloaded file, but they can’t change the hash displayed on the page. Where did it come from?

Then this, to me, is saying that hashes are better. Now how does that adversary change the THE TOR PROJECT, INC. signature and the countersignature from Digicert in that replaced download? Or do you mean the GPG signature when you said hashes. In this scenario the adversary could also hack the asc file downloaded.

> [@Preetor681](#):
>
> An adversary has root access to TorProject servers and can replace files with malicious ones, but can’t change the displayed hash on the page?

Agreed this adversary with root access could do anything.

---

<div class="post-metadata">

### Author: ![Lind](https://forum.torproject.org/user_avatar/forum.torproject.org/lind/32/7525_2.png) [@Lind](https://forum.torproject.org/u/Lind)
#### Post date: [February 17, 2026, 2:47pm UTC](https://forum.torproject.org/t/download-page-should-include-sha256-hash-values/21230/6 "2026-02-17T14:47:10Z")

</div>

> [@BobbyB](#):
>
> Then this, to me, is saying that hashes are better. Now how does that adversary change the THE TOR PROJECT, INC. signature and the countersignature from Digicert in that replaced download?

That person is talking about the original post,i.e. SHA256 (or some other) hashes, not GPG or Windows binary signature.

---

<div class="post-metadata">

### Author: ![compis](https://forum.torproject.org/letter_avatar_proxy/v4/letter/c/9d8465/32.png) [@compis](https://forum.torproject.org/u/compis)
#### Post date: [February 25, 2026, 10:23am UTC](https://forum.torproject.org/t/download-page-should-include-sha256-hash-values/21230/7 "2026-02-25T10:23:39Z")

</div>

A hash is simplier when displayed on the site and requires less effort to verify the app, you also have a backup from sites like virustotal to confirm the sha256 hash.

Since the hash is a simple character based value there should be no reason not to display it in the download along with the Signature and allow the users to choose.

---

<div class="post-metadata">

### Author: ![Lind](https://forum.torproject.org/user_avatar/forum.torproject.org/lind/32/7525_2.png) [@Lind](https://forum.torproject.org/u/Lind)
#### Post date: [February 25, 2026, 12:29pm UTC](https://forum.torproject.org/t/download-page-should-include-sha256-hash-values/21230/8 "2026-02-25T12:29:03Z")

</div>

A hash on website is also less capable of delivering desired security guarantee. Unless Tor Project website is redirecting users to mirror sites to download files (which it didn’t), the correctness of hashes and the integrity of the downloaded files relies on the same safety measure.

For the sake of clarity, let’s assume you went to Tor Project’s website to download the Browser…

1. If you’re served the correct hash, your TLS connection with Tor Project’s server is intact, and Tor Project’s server isn’t intruded by attackers, so you most likely already have downloaded genuine, unmodified Tor Browser.
2. If you’re served a modified Tor Browser, laced with malware, either Tor Project’s web servers have already been breached, or you’re suffering from a targetted MITM attack and the attackers somehow got a valid TLS certificate to impersonate Tor Project webservers. Either way, it’s unlikely you would be served a correct hash after downloading the mal-tor browser (it would be a rookie mistate NOT to do so, since attacker already has the capability to replace the correct hash either way).

---

<div class="post-metadata">

### Author: ![Lind](https://forum.torproject.org/user_avatar/forum.torproject.org/lind/32/7525_2.png) [@Lind](https://forum.torproject.org/u/Lind)
#### Post date: [February 25, 2026, 12:29pm UTC](https://forum.torproject.org/t/download-page-should-include-sha256-hash-values/21230/9 "2026-02-25T12:29:17Z")

</div>

> [@BobbyB](#):
>
> Not everyone has the utilities (or wants the utilities) to check the sig using the asc file like I do.

But on Windows, you also need specialized utility to obtain the hash of a file.  
Unless you use the PowerShell cmdlet `Get-FileHash`, but come on, how many Windows users even know PowerShell?

---

<div class="post-metadata">

### Author: ![BobbyB](https://forum.torproject.org/user_avatar/forum.torproject.org/bobbyb/32/5456_2.png) [@BobbyB](https://forum.torproject.org/u/BobbyB)
#### Post date: [February 25, 2026, 3:09pm UTC](https://forum.torproject.org/t/download-page-should-include-sha256-hash-values/21230/10 "2026-02-25T15:09:32Z")

</div>

> [@Lind](#):
>
> But on Windows, you also need specialized utility to obtain the hash of a file.

Not that specialized. 7-Zip portable is all you need but you don’t even need that. If you right click the file and go into properties you can check the digital signature of the file. I did an experiment. I changed a lowercase letter in the executable to an uppercase ( a 1 bit change) and it did not verify using the right click method.

> [@Lind](#):
>
> but come on, how many Windows users even know PowerShell?

Next to none. I know it’s there but never used it. Weird language. Wonder who invented it.  
Most Windows user never use the command line or even know it exists.

---

<div class="post-metadata">

### Author: ![Lind](https://forum.torproject.org/user_avatar/forum.torproject.org/lind/32/7525_2.png) [@Lind](https://forum.torproject.org/u/Lind)
#### Post date: [February 26, 2026, 3:10am UTC](https://forum.torproject.org/t/download-page-should-include-sha256-hash-values/21230/11 "2026-02-26T03:10:12Z")

</div>

> [@BobbyB](#):
>
> 7-Zip portable is all you need

Still more software to be installed. Also given that Windows 11 can now compress to 7z and tar and open rar archives, I’d expect less people feeling the need to install 7-Zip on their computers, and use of 7-Zip is limited to calculate file hashes for casual users that don’t need to deal with all kinds of archive files.

> [@BobbyB](#):
>
> digital signature

**signatures**. This is the second time I remind you that we’re talking about adding hash value of files to the website here, not digital signatures, neither PGP signatures nor Windows binary signatures.

Sure it’s probably good to add a guide about how to use Windows binary signatures to verify Tor Browsers, but please open another thread to discuss that matter. This is off topic for this thread.
