Measured on a real homelab
On our homelab it uses about 14 MB of RAM on a normal hour, 19 MB in busy hours, and peaked at 94 MB over the last 30 days (small library: 376 tracks (2.1 GB)).
The project publishes no RAM or CPU figure (only that ffmpeg is required and ARM boards like the Raspberry Pi are supported).
Sources
Navidrome publishes no official RAM figure. In the resource calculator we use 128 MB minimum and 256 MB recommended, derived from this measurement on a small library (process memory peaked at 94 MB; 123 MB counting disk cache): with a large collection, plan for more.
Our rule of thumb: plan for the highest process memory recorded plus 25% headroom: 128 MB. Our workload is small — scale it with the factors above before buying hardware.
Combine it with other apps in the resource calculator →
Build it: Rent a VPS · Buy a mini-PC
Memory per container from Netdata on the same server, summed across the whole stack (app, database and cache) hour by hour. The main figures are the processes' own memory (anonymous memory: heap, stacks). The docker stats figure is higher because it also counts active disk cache — which Linux gives back when it needs it — and shared memory such as PostgreSQL's shared buffers. Hours where the main container had no data are left out, not counted as zero. Re-measured with a script, not by hand.