How to Find Files on Linux: find, locate, fd and the Modern Toolkit

Linux doesn't have one file search — it has a toolbox, and knowing which tool to grab is the whole game. The classic answer to "linux find files" is the find command, and it's still the most capable; but the modern stack — fd for names, ripgrep for contents, fzf for fuzzy everything — is faster to type, faster to run, and kinder to your eyes. Add plocate for instant lookups, your desktop's built-in indexer, and Recoll when you need real full-text search over PDFs and documents, and nothing on your disks stays lost.

Every command below is copy-paste ready. This page is the hub for Linux search on this site; the guides it links go deeper on each technique.

find walks the file system live, which means it's always accurate and needs no index — and it can filter on practically any file attribute, then act on what it finds. The patterns you'll actually use:

# Every PDF anywhere under your home directory
find ~ -name '*.pdf'

# Same, but case-insensitive (catches .PDF, .Pdf)
find ~ -iname '*.pdf'

# Files over 100 MB — the "where did my disk go" query
find ~ -size +100M

# Files modified in the last 7 days
find ~ -mtime -7

# Combine: big videos changed this week, in one folder tree
find ~/Videos -iname '*.mkv' -size +500M -mtime -7

# Act on results: long-list every match
find ~ -name '*.log' -size +50M -exec ls -lh {} \;

# Delete with confirmation (safer than -delete)
find /tmp -name '*.tmp' -ok rm {} \;

Three details worth knowing: quote your patterns ('*.pdf') so the shell doesn't expand them early; -mtime -7 means "less than 7 days ago" while -mtime +7 means "more than"; and 2>/dev/null on the end silences the permission-denied noise when searching from /. For date-based hunting in depth — including the equivalents on Windows and macOS — see our guide to finding recently modified files, and for space audits, finding large files.

fd — find with sane defaults. fd does what find does for the common cases, with a fraction of the typing: smart-case matching, colored output, parallel traversal, and it skips hidden files and anything in your .gitignore by default (use -H and -I to bring them back).

# Everything matching "report" under the current directory
fd report

# By extension, under a specific folder
fd -e pdf . ~/Documents

# Changed in the last day, over 10 MB
fd --changed-within 1d --size +10m

# Run a command on every match, in parallel
fd -e jpg -x convert {} {.}.webp

fzf — fuzzy-find anything, interactively. fzf turns any list into a live, type-to-filter picker. With its shell bindings installed, Ctrl+T fuzzy-finds a file path into your current command line and Ctrl+R replaces your shell history search with something dramatically better. Type fragments — invrep24 finds invoices/report-2024.pdf.

# Pipe anything into an interactive picker
fd -e md | fzf

# Open the picked file in your editor
nvim "$(fzf)"

ripgrep — contents, not names. When the question is "which file contains this text," rg is the answer: recursive, gitignore-aware, and in our testing the fastest content grep available on any platform.

# Case-insensitive phrase search through a folder tree
rg -i "quarterly forecast" ~/Documents

# Only in Python files; show 2 lines of context
rg -t py "def main" -C 2

# Literal string (no regex interpretation)
rg -F "total: $99.95"

# Filenames only, including files rg would normally skip
rg -l --no-ignore "TODO"

rg takes full regular expressions, which is where it really earns its keep — our regex file search guide has ten copy-paste patterns that work in ripgrep, grepWin and most GUI tools. The caveat: like all grep-family tools, ripgrep reads plain text — it can't see inside PDFs or office documents. For those, skip ahead to Recoll.

GNOME Files + Tracker. GNOME's search (press the Super key and type, or Ctrl+F in Files) is backed by Tracker (now "TinySPARQL"), which indexes filenames and the contents of common document formats in your home directory. It's quietly decent: search from Files, filter by type and date from the dropdown, and results include content matches. Control what gets indexed under Settings → Search.

KDE Dolphin + Baloo. Dolphin's search box (Ctrl+F) rides on Baloo, KDE's indexer — famous mostly for the times it eats CPU. To tame it: System Settings → Search → File Search, switch from "Also index file content" to filename-only indexing (vastly lighter), and exclude build folders, VM images and caches. From a terminal, balooctl status shows progress, balooctl purge && balooctl enable rebuilds a corrupt index, and balooctl disable ends the relationship.

Desktop-agnostic GUIs. Catfish is a lightweight search window (a XFCE staple, runs anywhere) that wraps locate and live scanning with date/type filters — handy when you want clickable results without a heavyweight indexer. ANGRYsearch is the closest thing Linux has to the Windows tool Everything: it databases every filename and filters as you type, with the same freshness caveat as locate. If you've used Everything on Windows, ANGRYsearch will feel immediately familiar — and our free file search tools roundup puts the cross-platform picture together.

One more GUI-era problem worth naming: searching network shares. Whether it's a Samba server or a NAS, indexing rarely reaches across the wire — the options and trade-offs are covered in our network drive search guide.