LANDR is an online and in-DAW mastering service. It analyses a finished stereo mix and applies EQ, compression, stereo widening and limiting automatically, so you get a master in about a minute instead of the days a human engineer takes. You pick a style and a loudness level, preview the result, and download it. Masters you can download are only on paid LANDR Studio plans. The cheaper plans give you unlimited MP3 masters but a capped number of WAV files. LANDR hasn’t changed mastering standards. Loudness targets are still set by streaming platforms and standards bodies. What LANDR changed is the price of reaching those targets. For demos, content audio and fast singles it’s good enough. It can’t repair a bad mix, and on an important release it doesn’t replace a good human engineer.
How does LANDR’s AI mastering actually work?
You upload a stereo mix and LANDR analyses its frequency balance, dynamics and loudness. It then runs a mastering chain tuned to that analysis. You can change the style, the loudness, and a few tone controls, then compare the master with your original mix.
The same engine also runs as a plugin inside your DAW. The LANDR Mastering Plugin comes in SE and PRO editions. It has AI analysis, mastering styles, EQ, presence, a de-esser, stereo width, dynamics, gain match and loudness metering. It runs in VST3, AU and AAX on macOS and Windows. Because you’re mastering inside the session, you can change the mix and hear the master update straight away, which the browser upload can’t do.
What does LANDR cost?
Mastering comes as part of the LANDR Studio subscriptions. According to the LANDR mastering pricing page, Studio Standard includes unlimited MP3 masters, 36 WAV masters and the Mastering Plugin SE. Studio Pro adds unlimited HD WAV masters and the PRO plugin. Monthly prices for the tiers fall between about $10 and $40. Paying yearly cuts that a lot, and LANDR runs promotions often, so check the current price before you sign up.
The WAV cap is the limit that matters. Distributors and video editors need lossless files, and an MP3 master is no use for a release. If you master more than three tracks a month, you’ll either run out of WAVs on Standard or need Pro.
How loud should a master be for streaming?
There isn’t one universal number, but the main targets are well documented:
| Destination | Reference level | Source |
|---|---|---|
| Spotify (default playback) | −14 LUFS integrated. True peak below −1 dBTP, or below −2 dBTP if the master is louder than −14 LUFS | Spotify for Artists |
| Streaming music in general (AES recommendation) | About −16 LUFS for music, −18 LUFS for speech | Summary of AES TD1008 |
| European broadcast | −23 LUFS ±0.5 LU | EBU R 128 |
Spotify measures every track with the ITU-R BS.1770 method and turns it up or down at playback. Premium listeners can also choose Loud (−11 LUFS) or Quiet (−19 LUFS). The web player and third-party devices such as TVs and speakers don’t normalise at all. A master pushed to −8 LUFS gets turned down by about 6 dB on the default setting. All you keep from that extra loudness is the distortion it cost. Choose LANDR’s “Low” or “Medium” loudness unless your genre really needs to be dense.
Where does AI mastering fall short?
We use automated mastering a lot on music beds, social cuts and demo tracks for client approvals. These are the same failures we see again and again:
- It makes mix problems louder. A muddy 200–300 Hz range, a vocal that’s too quiet or a boomy kick all come out louder and more obvious. Fix the mix first.
- Harsh highs and sibilance. Bright styles can make esses and cymbals harsh. Use the de-esser or a warmer style.
- Peaks after encoding. Loud masters with a true peak close to 0 dBTP can clip once they’re converted to AAC or Ogg. Check true peak against Spotify’s −1 or −2 dBTP guidance.
- Albums don’t hang together. Each track is processed on its own, so levels and tone can drift from track to track. Use LANDR’s album mastering or a human engineer for a full record.
- No second opinion. A mastering engineer will tell you the bass is wrong before they master the track. The software just masters what you give it.
When should you use LANDR, and when should you hire a mastering engineer?
Use LANDR for demos, rough-mix references, podcast music, beds for video and social content, and singles on a small budget. In those cases speed and price matter more than the last 5% of quality. It’s also a useful check: if the AI master sounds wrong, the mix usually has a problem.
Hire an engineer for albums, vinyl, and paid sync or label releases. Human mastering costs anywhere from under $50 to several hundred dollars per track, depending on the engineer’s credits and how many revisions are included. That pays for listening on calibrated monitors, fixing problems in the mix, consistency across an album and properly prepared delivery files.
Does mastering for video differ from mastering for streaming?
Yes, and many people get this wrong. A track mastered for Spotify at about −14 LUFS is far louder than a broadcast deliverable. Broadcast programmes are measured as a whole, at −23 LUFS under EBU R 128 or −24 LKFS under the US ATSC recommendation. If you use the music master as the TV mix, it will be rejected at QC.
On our productions, the music master is only a source file. The final mix is done in the edit, balanced against dialogue, and measured to whatever level the delivery spec asks for. AI mastering saves time on the music. It doesn’t replace mixing to the delivery spec.
Next step: open your mix on LANDR’s free preview and set loudness to Medium. Compare the master with the original at matched volume using the gain-match option. If the AI master only sounds better because it’s louder, fix the mix first. If it holds up at matched volume, buy the plan that includes enough WAV masters for your release schedule.





