How to set up captcha verification on your Discord server

Why verify new members

Without verification, an account that joins can see everything and post everywhere from the first second. That is exactly what spam bots, raids and banned members coming back on a new account are counting on. Verification won’t stop everything, but it puts a delay between joining and the first message

Discord’s verification level and onboarding

Start with the native tools in your server settings. They cost nothing and apply before any bot even sees the member

• Verification level

  Depending on the level, it requires a verified email, an account older than 5 minutes, being on the server for more than 10 minutes, or a verified phone

• Onboarding

  On a Community server, Discord asks newcomers questions, gives them roles based on their answers and can have them accept the rules before they post

• Where they stop

  Both tools look at the account, not the person behind it. A bot with a verified email gets through, and nothing ties a new account to a member you banned

Kotbo’s three kinds of verification

Kotbo offers three paths. Keep one or stack them, depending on what your server deals with

• Captcha on join

  The newcomer gets an unverified role that only sees the verification channel. Kotbo posts an image there with a 6-character distorted code to type back. In voice mode, the code is read out in a voice channel opened for that member alone

• The rules button

  On the dashboard’s Rules page, turn on the accept button: it gives a role to whoever clicks it. The hide other channels option keeps everything else out of sight until they accept

• Identity verification

  The member signs in with Discord on a Kotbo web page. Kotbo compares the account, IP address and device fingerprint with past verifications on the server to spot an alt account

  Read the alt account article

Setting up a Discord captcha bot step by step

The captcha is the quickest to set up. Plan a few minutes and a second account to test with

1. 1

   Create the unverified role

   Create a role, say Unverified, that can only access the verification channel. Put it below Kotbo’s role in the role list

2. 2

   Prepare the verification channel

   A text channel where that role can read and post. Other members don’t need to see it

3. 3

   Turn the captcha on

   Run /protection captcha activer with the salon and role options. The delai and tentatives options set the time limit and number of tries, 10 minutes and 3 tries by default

4. 4

   Choose what happens on failure

   The sanction option decides what happens after the failed tries or once time runs out: a kick by default, or a ban

5. 5

   Give a role to those who pass

   On the dashboard’s Anti-raid page, captcha tab, pick the verified role. Kotbo grants it on success, then removes the unverified role

Verification commands

The captcha runs on join. Identity verification can also be requested later, for one member. Kotbo’s /protection subcommands and options are typed in French

/protection captcha activer

Turns the captcha on with the channel and the unverified role. Options: delai, tentatives, sanction, and salon-logs to keep a record of passes and failures

/protection captcha desactiver

Turns the captcha off. Newcomers join without a challenge again

/request-verification

Asks a member for an identity check, with the membre and raison options. They are timed out until they verify and get the link by DM, or in a private thread if their DMs are closed

Configuring identity verification

It lives on the dashboard under Suspicious accounts, Verification tab, and on the Rules page too. The settings that matter

• Delivery: an automatic DM on join, or a message with a button posted in your verification channel
• Level: Low reads only the Discord identity, Medium adds the email, High adds connected accounts and the server list
• When an alt is found: alert the staff, who confirm it or mark a false positive, or link the accounts automatically
• IP address and device fingerprint are saved by default, and each one can be turned off on its own
• A warning threshold can start the verification by itself when a member reaches it