Discord has no built-in recording feature. That's not an oversight - it's a deliberate design decision, partly because recording someone without their knowledge raises serious legal and ethical issues that Discord doesn't want to be in the middle of. The absence of a record button means you need an external tool, and your choice of tool determines how the recording sounds, how much it costs, and how much setup it requires.
There are three practical approaches: screen recording software running on your computer, a dedicated audio recording bot joining your voice channel, or a standalone audio recorder capturing your system output. Here's how each one works.

The fastest method: use OBS Studio (free) to record your screen and system audio while in a Discord call. Open OBS, add a Desktop Audio capture source, start recording, join your Discord voice channel, and OBS captures everything - your voice through the mic and others' voices through your speakers. For audio-only recording without video, Craig bot joins your voice channel and records each participant to a separate audio track automatically.
This guide is designed for readers who want to run a voice, video, screen sharing, streaming, or recording session with fewer technical surprises. It adds value beyond the basic menu path by explaining which device, permission, audio source, video quality, and privacy boundary should be set before the session starts. That matters because live features tend to fail at the worst moment when microphone access, screen permissions, bandwidth, or participant expectations were not tested.
The practical goal is to leave you with a result you can verify, not just a menu path you followed. It also keeps the limits visible: this guide does not pretend to offer perfect quality on every connection, secret recording, or a way to bypass platform and local consent rules. It shows the reliable path, the trade-offs, and the checks that help you avoid repeating the same mistake.
Recording isn't something most people do casually - there's usually a specific reason that makes it worth the setup.
You're running a podcast, interview, or collaborative project where the conversation is the content. Recording the call is the whole point of the session. You need clean audio from each participant, ideally on separate tracks so you can edit independently later. A bot like Craig handles this automatically - it joins the channel, records everyone on individual tracks, and sends you a download link when you're done. The alternative is recording the mixed output, which is harder to edit and sounds worse when one person's audio is uneven.
You're in a meeting or planning session where decisions and action items are made verbally, and you want a record to reference later. Teams that coordinate primarily through Discord voice channels often talk through important decisions that never make it into text. Recording these sessions - with everyone's knowledge and consent - gives you something to go back to when "didn't we agree on X?" becomes a question. Screen recording software works well here since you don't need the production quality of a podcast.
You're streaming, creating a tutorial, or capturing commentary alongside gameplay, and your Discord voice conversation is part of the content. In this case you're almost certainly already running OBS or similar software, and adding Discord audio to your recording setup is a configuration step rather than a separate tool. The challenge here is getting clean audio from both your microphone and Discord's output in a way that doesn't bleed or feedback - which the Advanced Tips section covers.
Recording a conversation without the knowledge and consent of all participants is illegal in many jurisdictions, and Discord's Terms of Service require you to comply with applicable laws. The legal standard varies: some places require only one-party consent (you are a participant, so your consent is sufficient), others require all-party consent. Before you record anything involving other people, tell them you're recording. This isn't just a legal precaution - it's basic respect. If someone objects, don't record.
This note applies regardless of method. The tool doesn't change the legal or ethical obligation.
OBS Studio is free, runs on Windows, Mac, and Linux, and captures both video and audio from your computer. It's more setup than the other methods but gives you the most control over what's recorded and how.
Go to obsproject.com and download OBS Studio for your operating system. Install and open it. The first time you launch, the Auto-Configuration Wizard offers to optimize settings - run it, and choose "Optimize just for recording, I will not be streaming."
In OBS, look at the Audio Mixer panel at the bottom. You should see two channels: Desktop Audio (system sound, which includes Discord's output) and Mic/Aux (your microphone). Click the gear icon next to Desktop Audio and select Properties to confirm it's capturing your system's default audio output. Do the same for Mic/Aux to confirm it's using your microphone. Both need to be active and showing levels when you speak and when Discord audio plays.
Go to Settings -> Output -> Recording. Choose a recording path where you want files saved. Set the format to MKV (safer than MP4 - if a recording stops unexpectedly, MKV files are recoverable while MP4 files may be corrupted). Set the recording quality to High Quality, Medium File Size for a good balance. Click OK.
Click Start Recording in OBS. Then join your Discord voice channel or call. OBS runs in the background capturing everything - your voice through the mic, other participants' voices through your speakers. When the call ends, click Stop Recording and find the file at the path you set.
Craig is a purpose-built Discord recording bot that joins your voice channel and records each participant on a separate audio track. It's the standard recommendation for podcasts, interviews, and any recording where you'll be editing the audio afterward.
Go to craig.chat and click Invite Craig. Authorize the bot on your server. Craig appears in your member list when added successfully.
Join the voice channel you want to record. Craig needs you to be in a voice channel before it will start recording there.
In any text channel, type the command :craig:, join. Craig will join the voice channel you're currently in and confirm in the text channel that recording has started. All participants in the channel are now being recorded on individual tracks.
When the call ends, type :craig:, leave in the text channel. Craig will leave the channel and send you a DM with a download link. The link gives you access to individual audio tracks for each participant - separate files you can mix and edit independently.
Click the download link in Craig's DM. You'll get a compressed archive containing separate audio files labeled by username. Import these into any audio editor (Audacity is free) to mix, cut, and export your final recording.
For a quick, no-bot, no-video recording of just the audio from a Discord call, you can use Audacity or your system's built-in audio recorder to capture what's playing through your speakers.
Download Audacity from audacityteam.org if you don't have it. Open it and go to Edit -> Preferences -> Devices. Under Recording, set the device to Stereo Mix or What U Hear (the naming varies by system). If neither option appears, you may need to enable it in your Windows Sound settings under Recording devices (right-click -> Show Disabled Devices, then enable Stereo Mix).
Click the Record button in Audacity, then join your Discord call. Audacity captures everything playing through your system - Discord's audio output. Your own voice through the microphone won't be captured this way unless you add a separate microphone input track in Audacity.
When the call ends, click Stop in Audacity. Go to File -> Export -> Export as MP3 (or WAV for lossless quality) to save the file.
OBS gives you a video file that captures whatever is on your screen alongside the audio - useful if you want a complete record of the session or plan to post the recording with visual context. The audio from OBS is a mixed recording: everyone's voices combined into a single track, which limits how much you can do in editing.
Craig produces individual tracks per participant, which is a fundamentally different starting point for editing. If one person's microphone is too loud, you adjust just their track. If there's a cough or interruption, you mute that moment on one track without affecting others. For anything that will be published or seriously edited, individual tracks are worth the marginal extra setup.
Audacity's Stereo Mix method is the simplest but least flexible. You get a single audio track of what played through your speakers - clean and useful for reference recordings, but not ideal for editing-heavy work.
All three methods produce local files that you control - nothing is stored on Discord's servers, nothing is uploaded automatically, and you decide who sees the recording.
Set Discord to output to a virtual audio cable to record participants' voices separately from your own mic. This is a Windows-specific setup using a free tool like VB-Audio VoiceMeeter or Virtual Audio Cable. Route Discord's output to a virtual device, then capture that device separately from your microphone in OBS or Audacity. The result is two separate audio sources - your mic and Discord's output - that you can adjust independently in post-production. This is the desktop alternative to Craig's multi-track approach.
Disable Discord's noise suppression before recording to preserve audio quality. Discord's noise suppression processes audio in real time, which can introduce subtle artifacts and reduce the naturalness of voices. For a recording you'll be editing, you generally want the raw audio rather than processed audio - you can apply your own noise reduction in post-production with better results. Turn off noise suppression in User Settings -> Voice & Video -> Advanced before recording sessions where quality matters.
Record a short test clip and play it back before the actual session starts. This takes two minutes and catches the most common recording failures: wrong input device selected, microphone not capturing, Discord audio not being recorded, levels too low or clipping. Finding out your recording setup wasn't working after a 90-minute session is one of the most avoidable frustrations in content creation. Test first, every time.
Use Audacity's Noise Reduction effect in post-production to clean up any background noise. Record 5-10 seconds of silence (just ambient room noise, no one speaking) at the start of your session. In Audacity, select that silent portion, go to Effect -> Noise Reduction -> Get Noise Profile, then select your entire recording and apply Noise Reduction. Audacity uses the noise profile from your silent sample to filter out that consistent background noise across the whole recording. It's a significant quality improvement for recordings made in less-than-ideal acoustic environments.
Every method described here records from your perspective - what your computer hears. If a participant has a poor microphone, a bad connection, or high background noise, that's what gets recorded. No amount of post-production fully fixes audio that was captured badly at the source. If recording quality matters, ask participants to use a headset rather than laptop speakers, join from a quiet location, and use a stable internet connection. The best recording setup can't compensate for a participant phoning in from a cafe on bluetooth earbuds.
Discord's processing also means that what you record isn't always what the other person's microphone actually captured. Discord's voice processing - echo cancellation, noise suppression, automatic gain control - runs on both ends of the call and changes how audio sounds before it reaches you. If you want truly unprocessed audio from participants, Craig with its individual track recording is the closest option, but even Craig records the processed audio that Discord transmits, not the raw microphone input on the other end.
Storage and file size are practical considerations that the original recording enthusiasm often overlooks. A two-hour OBS recording at high quality settings can be 10-30GB depending on your settings. Craig's individual tracks are smaller but multiply per participant. Make sure you have enough disk space before a long recording session, and have a plan for where the files go afterward.
OBS is recording but there's no audio from other Discord participants in the playback - only my microphone. What's wrong? The Desktop Audio source in OBS isn't capturing Discord's output. The most common cause is that Discord is routing its audio to a different device than what OBS's Desktop Audio is monitoring. Go to your Windows Sound settings and check which device is set as the Default Playback device - this is what Desktop Audio captures. Then check Discord's Voice & Video settings and confirm that its Output Device matches that same default. When they match, OBS will capture Discord's audio. If they're different, either change Discord's output to Default or configure OBS's Desktop Audio to capture the specific device Discord is using.
Craig recorded the call but some participants are missing from the individual track files - why? Craig records participants who are actively in the voice channel when recording starts and who speak during the session. Participants who join after recording has started may not be included, or their files may appear but contain only the audio from after they joined. Craig also has a file size limit on free recordings - very long sessions may be truncated. Check the download archive for a text file that lists which participants were captured and any warnings about file limits.
Is it possible to record a Discord call without the other person knowing, and would Discord detect this? Technically possible with screen recording software running silently in the background, yes. Discord has no mechanism to detect that you're recording and no way to alert other participants about it. But as noted before, recording without consent is illegal in many places and violates basic trust. Discord itself prohibits using the platform in ways that violate local laws. This guide doesn't encourage or endorse secret recording - it's here because people ask, and the honest answer is that Discord can't stop it while the law and ethics both say you shouldn't do it.
My Craig recording has a strange clicking or popping sound throughout - what causes this? This is almost always a sample rate mismatch between Craig's recording settings and your computer's audio output settings. Craig records at a specific sample rate, and if your system's audio device is configured to a different sample rate, the mismatch creates periodic artifacts in the recording. In Windows Sound settings, right-click your playback device -> Properties -> Advanced, and set the sample rate to 48000 Hz (48 kHz), which is Discord's native audio sample rate. Reconnect Craig and re-record to confirm the issue resolves.
If I use OBS to record a Discord call and post it online, do I own the recording? You own the recording as a technical file. What you can legally do with it depends on several factors: the consent of participants (recording and publishing without consent creates legal exposure in most jurisdictions), whether any copyrighted material played during the call (music, game audio, etc.), and the platform's terms of service where you upload. The recording itself doesn't carry automatic rights to publish - consent and copyright are separate questions that the recording method doesn't resolve.
A good finish looks like this: participants can hear, see, join, and leave as expected, and everyone knows what is being shared or recorded. Before moving on, do a small real-world test: start a short private test session, switch windows, mute and unmute, and check audio from another account or device.
If audio cuts out, the wrong screen is shared, the stream is laggy, or a recording is incomplete, the most likely explanation is that device permissions, hardware acceleration, network speed, server region, or the selected audio and video source is misconfigured. In that case, use the troubleshooting or limitation section above first, because repeating the same taps usually hides the real cause.
Read next: Use Discord voice channels | Change your Discord username | Stream games on Discord
OBS for general recording and Craig for podcast-quality multi-track audio cover the vast majority of use cases. The test clip before the actual session is the habit that prevents the most frustration - run it every time without exception. And tell people you're recording. It's a two-second disclosure that keeps everything above board and maintains the trust that makes the conversation worth recording in the first place.