How do I filter noise with `-fs`, `-fw`, `-fc`, and `-fmode`?

Filters drop responses; matchers keep them. Use `-fs` for size, `-fw` for word count, `-fl` for lines, `-fc` for status, and `-fr` for a regex. `-fmode and` or `or` (default `or`) sets how filters combine. `-ac` can calibrate size and word filters from sample responses.

How do `-recursion` and `-recursion-depth` work?

Enable `-recursion` with a URL that ends in FUZZ; ffuf queues each directory it finds as a new base. `-recursion-depth` caps how deep it goes (`0` means no depth cap). `-recursion-strategy` chooses `default` (redirect-based) or `greedy` (recurse on every match).

Can ffuf fuzz parameters and virtual hosts, not just directories?

Yes. Put FUZZ in the query string or a POST body (`-d`) to fuzz parameters, or in a Host header (`-H`) to enumerate vhosts against one IP. Use `-mode clusterbomb` or `pitchfork` with more than one wordlist to drive several FUZZ positions.

How do I control threads and request rate?

`-t` sets concurrent threads (default 40). `-rate` caps requests per second (`0` means no cap). `-p` adds a fixed or ranged delay between requests. Lower threads or rate, or add delay, to stay under WAF and rate-limit thresholds.

When should I use ffuf vs feroxbuster or ffuf-multi?

Use ffuf when you need one FUZZ engine for path and parameter jobs with matchers and filters you control. Prefer feroxbuster for recursive content discovery with opinionated defaults. Use ffuf-multi to run one wordlist across many URLs.

Updated Jul 14, 2026

Fuzzing

ffuf web fuzzer: directories, params, and vhosts

Q: What does `-mc all` do in ffuf?

`-mc` sets which HTTP status codes count as a match; default is 200,204,301,302,307,401,403. Pass `-mc all` to match every response, then remove noise with size or word filters like `-fs` and `-fw`. On Trickest, set the match-status-codes input.

A fast web fuzzer written in Go.

Discover hidden directories and files Find hidden parameters Enumerate virtual hosts

overview

What ffuf does

You already have a live host. Put FUZZ in the path for directories and files, in the query string or POST body (-d) for hidden parameters, or in a Host header (-H) for vhosts behind one IP. Each wordlist line fills FUZZ; ffuf records the response.

Speed only helps if you can cut noise. Keep responses with -mc (or -mc all), then drop decoys by size (-fs), words (-fw), lines (-fl), status (-fc), or regex (-fr). -ac can calibrate filter baselines from sample responses; -recursion with -recursion-depth walks into directories it finds; -mode clusterbomb or pitchfork drives several wordlists at once.

On Trickest, feed a target URL and a wordlist into the ffuf node; it writes a folder and a file. Enable -json or -of json so discovered paths, parameters, or hosts become structured output the next node can promote to a live table and probe.

source github.com/ffuf/ffuf

use cases

Where ffuf fits

Discover hidden directories and files

Fuzz a path with a content wordlist to surface admin panels, backups, and endpoints that are not linked anywhere.

Find hidden parameters

Fuzz the query string or POST body to reveal undocumented parameters worth testing for injection and access-control flaws.

Enumerate virtual hosts

Fuzz the Host header against one IP to find vhosts that do not appear in DNS.

Drive multiple injection points

Use clusterbomb or pitchfork mode with several wordlists to fuzz more than one FUZZ keyword in the same request.

reference

ffuf inputs and flags

68 inputs

Name	Type	Flag	Description
url	STRING	-u	Target URL, with FUZZ marking where each wordlist entry is injected.
wordlist	FILE	-w	Wordlist file path and optional keyword, separated by a colon (path:KEYWORD).
match-status-codes	STRING	-mc	Match HTTP status codes, or "all" (default 200,204,301,302,307,401,403).
filter-status-codes	STRING	-fc	Filter out responses by status code; comma-separated codes and ranges.
filter-response-size	STRING	-fs	Filter out responses by size; comma-separated sizes and ranges.
filter-words	STRING	-fw	Filter out responses by word count; comma-separated counts and ranges.
recursion	BOOLEAN	-recursion	Recurse into discovered directories (the URL must end in FUZZ).
extensions	STRING	-e	Comma-separated extensions that extend the FUZZ keyword (.php,.bak).
threads	STRING	-t	Number of concurrent threads (default 40).

Showing key inputs. ffuf exposes 68 inputs in total.

Full flag reference (68 inputs)

Name	Type	Flag	Description
url	STRING	-u	Target URL, with FUZZ where the wordlist is injected.
wordlist	FILE	-w	Wordlist file path and optional keyword, separated by a colon.
mode	STRING	-mode	Multi-wordlist mode: clusterbomb, pitchfork, sniper (default clusterbomb).
extensions	STRING	-e	Comma-separated list of extensions that extend the FUZZ keyword.
match-status-codes	STRING	-mc	Match HTTP status codes, or "all" (default 200,204,301,302,307,401,403).
match-response-size	STRING	-ms	Match HTTP response size.
match-words	STRING	-mw	Match a number of words in the response.
match-lines	STRING	-ml	Match a number of lines in the response.
match-regexp	STRING	-mr	Match a regexp against the response.
match-time	STRING	-mt	Match milliseconds to the first response byte, e.g. >100 or <100.
matcher-mode	STRING	-mmode	Matcher set operator: and, or (default or).
filter-status-codes	STRING	-fc	Filter HTTP status codes; comma-separated list of codes and ranges.
filter-response-size	STRING	-fs	Filter by response size; comma-separated list of sizes and ranges.
filter-words	STRING	-fw	Filter by word count; comma-separated list of counts and ranges.
filter-lines	STRING	-fl	Filter by line count; comma-separated list of counts and ranges.
filter-regexp	STRING	-fr	Filter by a regexp against the response.
filter-time	STRING	-ft	Filter by milliseconds to the first response byte, e.g. >100 or <100.
filter-mode	STRING	-fmode	Filter set operator: and, or (default or).
autocalibration	BOOLEAN	-ac	Automatically calibrate filtering options (default false).
host-autocalibration	BOOLEAN	-ach	Per-host autocalibration (default false).
custom-autocalibration	STRING	-acc	Custom auto-calibration string; can be repeated, implies -ac.
autocalibration-strategy	STRING	-acs	Custom auto-calibration strategy; can be repeated, implies -ac.
autocalibration-keyword	STRING	-ack	Autocalibration keyword (default FUZZ).
recursion	BOOLEAN	-recursion	Scan recursively; only FUZZ is supported and the URL must end in it (default false).
recursion-depth	STRING	-recursion-depth	Maximum recursion depth (default 0, unlimited).
recursion-strategy	STRING	-recursion-strategy	Recursion strategy: default (redirect-based) or greedy (default default).
threads	STRING	-t	Number of concurrent threads (default 40).
rate	STRING	-rate	Rate of requests per second (default 0, unlimited).
delay	STRING	-p	Seconds of delay between requests, or a range, e.g. 0.1 or 0.1-2.0.
timeout	STRING	-timeout	HTTP request timeout in seconds (default 10).
maxtime	STRING	-maxtime	Maximum running time in seconds for the entire process (default 0).
maxtime-job	STRING	-maxtime-job	Maximum running time in seconds per job (default 0).
header	STRING	-H	Header "Name: Value", separated by a colon.
header-file	FILE	-header-file	Headers "Name: Value", one per newline.
http-method	STRING	-X	HTTP method to use (default GET).
post-data	STRING	-d	POST data.
cookie-data	STRING	-b	Cookie data.
request	FILE	-request	File containing the raw HTTP request to fuzz.
request-proto	STRING	-request-proto	Protocol to use with a raw request (default https).
raw	BOOLEAN	-raw	Do not encode the URI (default false).
http2	BOOLEAN	-http2	Use the HTTP/2 protocol (default false).
sni	STRING	-sni	Target TLS SNI; does not support the FUZZ keyword.
follow-redirects	BOOLEAN	-r	Follow redirects (default false).
http-proxy-url	STRING	-x	Proxy URL (SOCKS5 or HTTP), e.g. http://127.0.0.1:8080.
replay-proxy	STRING	-replay-proxy	Replay matched requests through this proxy.
client-cert	FILE	-cc	Client cert for authentication; requires a client key.
client-key	FILE	-ck	Client key for authentication; requires a client certificate.
encoders	STRING	-enc	Encoders for keywords, e.g. FUZZ:urlencode b64encode.
input-cmd	STRING	-input-cmd	Command producing the input; requires -input-num and overrides -w.
input-num	STRING	-input-num	Number of inputs to test, used with -input-cmd (default 100).
input-shell	STRING	-input-shell	Shell used for running the input command.
ignore-wordlist-comments	BOOLEAN	-ic	Ignore wordlist comments (default false).
dirsearch-wordlist-compatibility-mode	BOOLEAN	-D	DirSearch wordlist compatibility mode; used with -e (default false).
json	BOOLEAN	-json	JSON output, printing newline-delimited JSON records (default false).
output-format	STRING	-of	Output file format: json, ejson, html, md, csv, ecsv (default json).
output-skip-empty-file	BOOLEAN	-or	Do not create the output file when there are no results (default false).
silent	BOOLEAN	-s	Silent mode; do not print additional information (default false).
verbose	BOOLEAN	-v	Verbose output, printing the full URL and any redirect location (default false).
color-output	BOOLEAN	-c	Colorize the output.
ignore-body	BOOLEAN	-ignore-body	Do not fetch the response content (default false).
dont-fetch-response-body	BOOLEAN	-ignore-body	Do not fetch the response content (default false).
stop-on-forbidden	BOOLEAN	-sf	Stop when over 95% of responses return 403 Forbidden (default false).
stop-on-spurious-errors	BOOLEAN	-se	Stop on spurious errors (default false).
stop-on-all-errors	BOOLEAN	-sa	Stop on all error cases; implies -sf and -se (default false).
scrapers	STRING	-scrapers	Active scraper groups (default all).
scraperfile	FILE	-scraperfile	Custom scraper file path.
search	STRING	-search	Search for a FFUFHASH payload from ffuf history.
config	FILE	-config	Load configuration from a file.

example

Run ffuf

ffuf · command

# fuzz for hidden paths, match every code, drop 4242-byte decoy responsesffuf -u https://example.com/FUZZ \  -w /usr/share/seclists/Discovery/Web-Content/raft-medium-directories.txt \  -mc all -fs 4242 -e .php,.html \  -recursion -recursion-depth 2 -t 40 \  -o results.json -of json

sample output

 :: Method           : GET :: URL              : https://example.com/FUZZ :: Wordlist         : FUZZ: raft-medium-directories.txt :: Matcher          : Response status: all :: Filter           : Response size: 4242________________________________________________ admin                   [Status: 301, Size: 178, Words: 6, Lines: 8, Duration: 44ms]login                   [Status: 200, Size: 2011, Words: 213, Lines: 61, Duration: 91ms].git/HEAD               [Status: 200, Size: 23, Words: 2, Lines: 1, Duration: 63ms]backup                  [Status: 403, Size: 279, Words: 20, Lines: 10, Duration: 38ms]:: Progress: [30000/30000] :: Job [1/1] :: 980 req/sec :: Duration: [0:00:31] :: Errors: 0 ::

guidance

Choosing ffuf

Use ffuf to discover content, parameters, or vhosts on a host you already know is live. Filter responses by status, size, or words to cut noise. To run one wordlist across many URLs, use ffuf-multi; for recursive defaults, feroxbuster.

feroxbuster

Recursive content discovery with opinionated defaults. ffuf gives you one engine and full matcher/filter control.

gobuster

Classic directory and DNS brute force. ffuf adds parameter fuzzing and richer filtering.

ffuf-multi

Same ffuf engine across many URLs in one run. Use plain ffuf for a single target and wordlist.

faq

ffuf questions

`-mc` sets which HTTP status codes count as a match; default is 200,204,301,302,307,401,403. Pass `-mc all` to match every response, then remove noise with size or word filters like `-fs` and `-fw`. On Trickest, set the match-status-codes input.

Run ffuf yourself

A target and a wordlist feed ffuf, which fuzzes the request and writes the discovered paths as a queryable output.

Open ffuf in the library Build a workflow with it

Facts on this page come from the live Trickest tool library.