1YOUTUBE-DL(1) YOUTUBE-DL(1)
2
6 youtube-dl - download videos from youtube.com or other video platforms
7
9 youtube-dl [OPTIONS] URL [URL...]
10
12 youtube-dl is a command-line program to download videos from
13 YouTube.com and a few more sites. It requires the Python interpreter,
14 version 2.6, 2.7, or 3.2+, and it is not platform specific. It should
15 work on your Unix box, on Windows or on macOS. It is released to the
16 public domain, which means you can modify it, redistribute it or use it
17 however you like.
18
20 -h, –help
21 Print this help text and exit
22 –version
24 Print program version and exit
25 -U, –update
27 Update this program to latest version. Make sure that you have
28 sufficient permissions (run with sudo if needed)
29 -i, –ignore-errors
31 Continue on download errors, for example to skip unavailable
32 videos in a playlist
33 –abort-on-error
35 Abort downloading of further videos (in the playlist or the com‐
36 mand line) if an error occurs
37 –dump-user-agent
39 Display the current browser identification
40 –list-extractors
42 List all supported extractors
43 –extractor-descriptions
45 Output descriptions of all supported extractors
46 –force-generic-extractor
48 Force extraction to use the generic extractor
49 –default-search PREFIX
51 Use this prefix for unqualified URLs. For example “gvsearch2:”
52 downloads two videos from google videos for youtube- dl “large
53 apple”. Use the value “auto” to let youtube-dl guess (“au‐
54 to_warning” to emit a warning when guessing). “error” just
55 throws an error. The default value “fixup_error” repairs broken
56 URLs, but emits an error if this is not possible instead of
57 searching.
58 –ignore-config
60 Do not read configuration files. When given in the global con‐
61 figuration file /etc/youtube-dl.conf: Do not read the user con‐
62 figuration in ~/.config/youtube-dl/config (%APPDATA%/youtube-
63 dl/config.txt on Windows)
64 –config-location PATH
66 Location of the configuration file; either the path to the con‐
67 fig or its containing directory.
68 –flat-playlist
70 Do not extract the videos of a playlist, only list them.
71 –mark-watched
73 Mark videos watched (YouTube only)
74 –no-mark-watched
76 Do not mark videos watched (YouTube only)
77 –no-color
79 Do not emit color codes in output
80 Network Options:
82 –proxy URL
83 Use the specified HTTP/HTTPS/SOCKS proxy. To enable SOCKS
84 proxy, specify a proper scheme. For example
85 socks5://127.0.0.1:1080/. Pass in an empty string (–proxy ““)
86 for direct connection
87 –socket-timeout SECONDS
89 Time to wait before giving up, in seconds
90 –source-address IP
92 Client-side IP address to bind to
93 -4, –force-ipv4
95 Make all connections via IPv4
96 -6, –force-ipv6
98 Make all connections via IPv6
99 Geo Restriction:
101 –geo-verification-proxy URL
102 Use this proxy to verify the IP address for some geo-restricted
103 sites. The default proxy specified by –proxy (or none, if the
104 option is not present) is used for the actual downloading.
105 –geo-bypass
107 Bypass geographic restriction via faking X-Forwarded-For HTTP
108 header
109 –no-geo-bypass
111 Do not bypass geographic restriction via faking X-Forwarded-For
112 HTTP header
113 –geo-bypass-country CODE
115 Force bypass geographic restriction with explicitly provided
116 two-letter ISO 3166-2 country code
117 –geo-bypass-ip-block IP_BLOCK
119 Force bypass geographic restriction with explicitly provided IP
120 block in CIDR notation
121 Video Selection:
123 –playlist-start NUMBER
124 Playlist video to start at (default is 1)
125 –playlist-end NUMBER
127 Playlist video to end at (default is last)
128 –playlist-items ITEM_SPEC
130 Playlist video items to download. Specify indices of the videos
131 in the playlist separated by commas like: “– playlist-items
132 1,2,5,8” if you want to download videos indexed 1, 2, 5, 8 in
133 the playlist. You can specify range: ” –playlist-items
134 1-3,7,10-13”, it will download the videos at index 1, 2, 3, 7,
135 10, 11, 12 and 13.
136 –match-title REGEX
138 Download only matching titles (case- insensitive regex or al‐
139 phanumeric sub- string)
140 –reject-title REGEX
142 Skip download for matching titles (case-insensitive regex or al‐
143 phanumeric sub-string)
144 –max-downloads NUMBER
146 Abort after downloading NUMBER files
147 –min-filesize SIZE
149 Do not download any videos smaller than SIZE (e.g. 50k or 44.6m)
150 –max-filesize SIZE
152 Do not download any videos larger than SIZE (e.g. 50k or 44.6m)
153 –date DATE
155 Download only videos uploaded in this date
156 –datebefore DATE
158 Download only videos uploaded on or before this date (i.e. in‐
159 clusive)
160 –dateafter DATE
162 Download only videos uploaded on or after this date (i.e. inclu‐
163 sive)
164 –min-views COUNT
166 Do not download any videos with less than COUNT views
167 –max-views COUNT
169 Do not download any videos with more than COUNT views
170 –match-filter FILTER
172 Generic video filter. Specify any key (see the “OUTPUT TEM‐
173 PLATE” for a list of available keys) to match if the key is
174 present, !key to check if the key is not present, key > NUMBER
175 (like “comment_count > 12”, also works with >=, <, <=, !=, =) to
176 compare against a number, key = `LITERAL' (like “uploader =
177 `Mike Smith'”, also works with !=) to match against a string
178 literal and & to require multiple matches. Values which are not
179 known are excluded unless you put a question mark (?) after the
180 operator. For example, to only match videos that have been
181 liked more than 100 times and disliked less than 50 times (or
182 the dislike functionality is not available at the given ser‐
183 vice), but who also have a description, use –match-filter
184 “like_count > 100 & dislike_count <? 50 & description” .
185 –no-playlist
187 Download only the video, if the URL refers to a video and a
188 playlist.
189 –yes-playlist
191 Download the playlist, if the URL refers to a video and a
192 playlist.
193 –age-limit YEARS
195 Download only videos suitable for the given age
196 –download-archive FILE
198 Download only videos not listed in the archive file. Record the
199 IDs of all downloaded videos in it.
200 –include-ads
202 Download advertisements as well (experimental)
203 Download Options:
205 -r, –limit-rate RATE
206 Maximum download rate in bytes per second (e.g. 50K or 4.2M)
207 -R, –retries RETRIES
209 Number of retries (default is 10), or “infinite”.
210 –fragment-retries RETRIES
212 Number of retries for a fragment (default is 10), or “infinite”
213 (DASH, hlsnative and ISM)
214 –skip-unavailable-fragments
216 Skip unavailable fragments (DASH, hlsnative and ISM)
217 –abort-on-unavailable-fragment
219 Abort downloading when some fragment is not available
220 –keep-fragments
222 Keep downloaded fragments on disk after downloading is finished;
223 fragments are erased by default
224 –buffer-size SIZE
226 Size of download buffer (e.g. 1024 or 16K) (default is 1024)
227 –no-resize-buffer
229 Do not automatically adjust the buffer size. By default, the
230 buffer size is automatically resized from an initial value of
231 SIZE.
232 –http-chunk-size SIZE
234 Size of a chunk for chunk-based HTTP downloading (e.g. 10485760
235 or 10M) (default is disabled). May be useful for bypassing
236 bandwidth throttling imposed by a webserver (experimental)
237 –playlist-reverse
239 Download playlist videos in reverse order
240 –playlist-random
242 Download playlist videos in random order
243 –xattr-set-filesize
245 Set file xattribute ytdl.filesize with expected file size
246 –hls-prefer-native
248 Use the native HLS downloader instead of ffmpeg
249 –hls-prefer-ffmpeg
251 Use ffmpeg instead of the native HLS downloader
252 –hls-use-mpegts
254 Use the mpegts container for HLS videos, allowing to play the
255 video while downloading (some players may not be able to play
256 it)
257 –external-downloader COMMAND
259 Use the specified external downloader. Currently supports
260 aria2c,aria2p,avconv ,axel,curl,ffmpeg,httpie,wget
261 –external-downloader-args ARGS
263 Give these arguments to the external downloader
264 Filesystem Options:
266 -a, –batch-file FILE
267 File containing URLs to download (`-' for stdin), one URL per
268 line. Lines starting with `#', `;' or `]' are considered as
269 comments and ignored.
270 –id Use only video ID in file name
272 -o, –output TEMPLATE
274 Output filename template, see the “OUTPUT TEMPLATE” for all the
275 info
276 –output-na-placeholder PLACEHOLDER
278 Placeholder value for unavailable meta fields in output filename
279 template (default is “NA”)
280 –autonumber-start NUMBER
282 Specify the start value for %(autonumber)s (default is 1)
283 –restrict-filenames
285 Restrict filenames to only ASCII characters, and avoid “&” and
286 spaces in filenames
287 -w, –no-overwrites
289 Do not overwrite files
290 -c, –continue
292 Force resume of partially downloaded files. By default,
293 youtube-dl will resume downloads if possible.
294 –no-continue
296 Do not resume partially downloaded files (restart from begin‐
297 ning)
298 –no-part
300 Do not use .part files - write directly into output file
301 –mtime Use the Last-modified header to set the file modification time
303 (default)
304 –no-mtime
306 Do not use the Last-modified header to set the file modification
307 time
308 –write-description
310 Write video description to a .description file
311 –write-info-json
313 Write video metadata to a .info.json file
314 –write-annotations
316 Write video annotations to a .annotations.xml file
317 –load-info-json FILE
319 JSON file containing the video information (created with the
320 “–write- info-json” option)
321 –cookies FILE
323 File to read cookies from and dump cookie jar in
324 –cache-dir DIR
326 Location in the filesystem where youtube-dl can store some down‐
327 loaded information permanently. By default
328 $XDG_CACHE_HOME/youtube-dl or ~/.cache/youtube-dl . At the mo‐
329 ment, only YouTube player files (for videos with obfuscated sig‐
330 natures) are cached, but that may change.
331 –no-cache-dir
333 Disable filesystem caching
334 –rm-cache-dir
336 Delete all filesystem cache files
337 Thumbnail Options:
339 –write-thumbnail
340 Write thumbnail image to disk
341 –write-all-thumbnails
343 Write all thumbnail image formats to disk
344 –list-thumbnails
346 Simulate and list all available thumbnail formats
347 Verbosity / Simulation Options:
349 -q, –quiet
350 Activate quiet mode
351 –no-warnings
353 Ignore warnings
354 -s, –simulate
356 Do not download the video and do not write anything to disk
357 –skip-download
359 Do not download the video
360 -g, –get-url
362 Simulate, quiet but print URL
363 -e, –get-title
365 Simulate, quiet but print title
366 –get-id
368 Simulate, quiet but print id
369 –get-thumbnail
371 Simulate, quiet but print thumbnail URL
372 –get-description
374 Simulate, quiet but print video description
375 –get-duration
377 Simulate, quiet but print video length
378 –get-filename
380 Simulate, quiet but print output filename
381 –get-format
383 Simulate, quiet but print output format
384 -j, –dump-json
386 Simulate, quiet but print JSON information. See the “OUTPUT
387 TEMPLATE” for a description of available keys.
388 -J, –dump-single-json
390 Simulate, quiet but print JSON information for each command-line
391 argument. If the URL refers to a playlist, dump the whole
392 playlist information in a single line.
393 –print-json
395 Be quiet and print the video information as JSON (video is still
396 being downloaded).
397 –newline
399 Output progress bar as new lines
400 –no-progress
402 Do not print progress bar
403 –console-title
405 Display progress in console titlebar
406 -v, –verbose
408 Print various debugging information
409 –dump-pages
411 Print downloaded pages encoded using base64 to debug problems
412 (very verbose)
413 –write-pages
415 Write downloaded intermediary pages to files in the current di‐
416 rectory to debug problems
417 –print-traffic
419 Display sent and read HTTP traffic
420 -C, –call-home
422 Contact the youtube-dl server for debugging
423 –no-call-home
425 Do NOT contact the youtube-dl server for debugging
426 Workarounds:
428 –encoding ENCODING
429 Force the specified encoding (experimental)
430 –no-check-certificate
432 Suppress HTTPS certificate validation
433 –prefer-insecure
435 Use an unencrypted connection to retrieve information about the
436 video. (Currently supported only for YouTube)
437 –user-agent UA
439 Specify a custom user agent
440 –referer URL
442 Specify a custom Referer: use if the video access is restricted
443 to one domain
444 –add-header FIELD:VALUE
446 Specify a custom HTTP header and its value, separated by a colon
447 `:'. You can use this option multiple times. NB Use –cookies
448 rather than adding a Cookie header if its contents may be sensi‐
449 tive; data from a Cookie header will be sent to all domains, not
450 just the one intended
451 –bidi-workaround
453 Work around terminals that lack bidirectional text support. Re‐
454 quires bidiv or fribidi executable in PATH
455 –sleep-interval SECONDS
457 Number of seconds to sleep before each download when used alone
458 or a lower bound of a range for randomized sleep before each
459 download (minimum possible number of seconds to sleep) when used
460 along with –max-sleep-interval.
461 –max-sleep-interval SECONDS
463 Upper bound of a range for randomized sleep before each download
464 (maximum possible number of seconds to sleep). Must only be
465 used along with –min- sleep-interval.
466 Video Format Options:
468 -f, –format FORMAT
469 Video format code, see the “FORMAT SELECTION” for all the info
470 –all-formats
472 Download all available video formats
473 –prefer-free-formats
475 Prefer free video formats unless a specific one is requested
476 -F, –list-formats
478 List all available formats of requested videos
479 –youtube-skip-dash-manifest
481 Do not download the DASH manifests and related data on YouTube
482 videos
483 –merge-output-format FORMAT
485 If a merge is required (e.g. bestvideo+bestaudio), output to
486 given container format. One of mkv, mp4, ogg, webm, flv. Ig‐
487 nored if no merge is required
488 Subtitle Options:
490 –write-sub
491 Write subtitle file
492 –write-auto-sub
494 Write automatically generated subtitle file (YouTube only)
495 –all-subs
497 Download all the available subtitles of the video
498 –list-subs
500 List all available subtitles for the video
501 –sub-format FORMAT
503 Subtitle format, accepts formats preference, for example: “srt”
504 or “ass/srt/best”
505 –sub-lang LANGS
507 Languages of the subtitles to download (optional) separated by
508 commas, use –list-subs for available language tags
509 Authentication Options:
511 -u, –username USERNAME
512 Login with this account ID
513 -p, –password PASSWORD
515 Account password. If this option is left out, youtube-dl will
516 ask interactively.
517 -2, –twofactor TWOFACTOR
519 Two-factor authentication code
520 -n, –netrc
522 Use .netrc authentication data
523 –video-password PASSWORD
525 Video password (vimeo, youku)
526 Adobe Pass Options:
528 –ap-mso MSO
529 Adobe Pass multiple-system operator (TV provider) identifier,
530 use –ap-list-mso for a list of available MSOs
531 –ap-username USERNAME
533 Multiple-system operator account login
534 –ap-password PASSWORD
536 Multiple-system operator account password. If this option is
537 left out, youtube-dl will ask interactively.
538 –ap-list-mso
540 List all supported multiple-system operators
541 Post-processing Options:
543 -x, –extract-audio
544 Convert video files to audio-only files (requires ffmpeg/avconv
545 and ffprobe/avprobe)
546 –audio-format FORMAT
548 Specify audio format: “best”, “aac”, “flac”, “mp3”, “m4a”,
549 “opus”, “vorbis”, or “wav”; “best” by default; No effect without
550 -x
551 –audio-quality QUALITY
553 Specify ffmpeg/avconv audio quality, insert a value between 0
554 (better) and 9 (worse) for VBR or a specific bitrate like 128K
555 (default 5)
556 –recode-video FORMAT
558 Encode the video to another format if necessary (currently sup‐
559 ported: mp4|flv|ogg|webm|mkv|avi)
560 –postprocessor-args ARGS
562 Give these arguments to the postprocessor (if postprocessing is
563 required)
564 -k, –keep-video
566 Keep the video file on disk after the post-processing; the video
567 is erased by default
568 –no-post-overwrites
570 Do not overwrite post-processed files; the post-processed files
571 are overwritten by default
572 –embed-subs
574 Embed subtitles in the video (only for mp4, webm and mkv videos)
575 –embed-thumbnail
577 Embed thumbnail in the audio as cover art
578 –add-metadata
580 Write metadata to the video file
581 –metadata-from-title FORMAT
583 Parse additional metadata like song title / artist from the
584 video title. The format syntax is the same as –output. Regular
585 expression with named capture groups may also be used. The
586 parsed parameters replace existing values. Example: –metadata-
587 from-title “%(artist)s - %(title)s” matches a title like “Cold‐
588 play - Paradise”. Example (regex): –metadata-from-title
589 “(?P.+?) - (?P
590 .+)”
591 –xattrs
593 Write metadata to the video file’s xattrs (using dublin core and
594 xdg standards)
595 –fixup POLICY
597 Automatically correct known faults of the file. One of never
598 (do nothing), warn (only emit a warning), detect_or_warn (the
599 default; fix file if we can, warn otherwise)
600 –prefer-avconv
602 Prefer avconv over ffmpeg for running the postprocessors
603 –prefer-ffmpeg
605 Prefer ffmpeg over avconv for running the postprocessors (de‐
606 fault)
607 –ffmpeg-location PATH
609 Location of the ffmpeg/avconv binary; either the path to the bi‐
610 nary or its containing directory.
611 –exec CMD
613 Execute a command on the file after downloading and post-pro‐
614 cessing, similar to find’s -exec syntax. Example: –exec `adb
615 push {} /sdcard/Music/ && rm {}'
616 –convert-subs FORMAT
618 Convert the subtitles to other format (currently supported:
619 srt|ass|vtt|lrc)
620
622 You can configure youtube-dl by placing any supported command line op‐
623 tion to a configuration file. On Linux and macOS, the system wide con‐
624 figuration file is located at /etc/youtube-dl.conf and the user wide
625 configuration file at ~/.config/youtube-dl/config. On Windows, the us‐
626 er wide configuration file locations are %APPDATA%\youtube-dl\con‐
627 fig.txt or C:\Users\<user name>\youtube-dl.conf. Note that by default
628 configuration file may not exist so you may need to create it yourself.
629 For example, with the following configuration file youtube-dl will al‐
631 ways extract the audio, not copy the mtime, use a proxy and save all
632 videos under Movies directory in your home directory:
633 # Lines starting with # are comments
635 # Always extract audio
637 -x
638 # Do not copy the mtime
640 --no-mtime
641 # Use this proxy
643 --proxy 127.0.0.1:3128
644 # Save all videos under Movies directory in your home directory
646 -o ~/Movies/%(title)s.%(ext)s
647 Note that options in configuration file are just the same options aka
649 switches used in regular command line calls thus there must be no
650 whitespace after - or --, e.g. -o or --proxy but not - o or -- proxy.
651 You can use --ignore-config if you want to disable the configuration
653 file for a particular youtube-dl run.
654 You can also use --config-location if you want to use custom configura‐
656 tion file for a particular youtube-dl run.
657 Authentication with .netrc file
659 You may also want to configure automatic credentials storage for ex‐
660 tractors that support authentication (by providing login and password
661 with --username and --password) in order not to pass credentials as
662 command line arguments on every youtube-dl execution and prevent track‐
663 ing plain text passwords in the shell command history. You can achieve
664 this using a .netrc file (https://stackoverflow.com/tags/.netrc/info)
665 on a per extractor basis. For that you will need to create a .netrc
666 file in your $HOME and restrict permissions to read/write by only you:
667 touch $HOME/.netrc
669 chmod a-rwx,u+rw $HOME/.netrc
670 After that you can add credentials for an extractor in the following
672 format, where extractor is the name of the extractor in lowercase:
673 machine <extractor> login <login> password <password>
675 For example:
677 machine youtube login myaccount@gmail.com password my_youtube_password
679 machine twitch login my_twitch_account_name password my_twitch_password
680 To activate authentication with the .netrc file you should pass --netrc
682 to youtube-dl or place it in the configuration file.
683 On Windows you may also need to setup the %HOME% environment variable
685 manually. For example:
686 set HOME=%USERPROFILE%
688
690 The -o option allows users to indicate a template for the output file
691 names.
692 tl;dr: navigate me to examples.
694 The basic usage is not to set any template arguments when downloading a
696 single file, like in youtube-dl -o funny_video.flv
697 "https://some/video". However, it may contain special sequences that
698 will be replaced when downloading each video. The special sequences
699 may be formatted according to python string formatting operations
700 (https://docs.python.org/2/library/stdtypes.html#string-formatting).
701 For example, %(NAME)s or %(NAME)05d. To clarify, that is a percent
702 symbol followed by a name in parentheses, followed by formatting opera‐
703 tions. Allowed names along with sequence type are:
704 • id (string): Video identifier
706 • title (string): Video title
708 • url (string): Video URL
710 • ext (string): Video filename extension
712 • alt_title (string): A secondary title of the video
714 • display_id (string): An alternative identifier for the video
716 • uploader (string): Full name of the video uploader
718 • license (string): License name the video is licensed under
720 • creator (string): The creator of the video
722 • release_date (string): The date (YYYYMMDD) when the video was re‐
724 leased
725 • timestamp (numeric): UNIX timestamp of the moment the video became
727 available
728 • upload_date (string): Video upload date (YYYYMMDD)
730 • uploader_id (string): Nickname or id of the video uploader
732 • channel (string): Full name of the channel the video is uploaded on
734 • channel_id (string): Id of the channel
736 • location (string): Physical location where the video was filmed
738 • duration (numeric): Length of the video in seconds
740 • view_count (numeric): How many users have watched the video on the
742 platform
743 • like_count (numeric): Number of positive ratings of the video
745 • dislike_count (numeric): Number of negative ratings of the video
747 • repost_count (numeric): Number of reposts of the video
749 • average_rating (numeric): Average rating give by users, the scale
751 used depends on the webpage
752 • comment_count (numeric): Number of comments on the video
754 • age_limit (numeric): Age restriction for the video (years)
756 • is_live (boolean): Whether this video is a live stream or a fixed-
758 length video
759 • start_time (numeric): Time in seconds where the reproduction should
761 start, as specified in the URL
762 • end_time (numeric): Time in seconds where the reproduction should
764 end, as specified in the URL
765 • format (string): A human-readable description of the format
767 • format_id (string): Format code specified by --format
769 • format_note (string): Additional info about the format
771 • width (numeric): Width of the video
773 • height (numeric): Height of the video
775 • resolution (string): Textual description of width and height
777 • tbr (numeric): Average bitrate of audio and video in KBit/s
779 • abr (numeric): Average audio bitrate in KBit/s
781 • acodec (string): Name of the audio codec in use
783 • asr (numeric): Audio sampling rate in Hertz
785 • vbr (numeric): Average video bitrate in KBit/s
787 • fps (numeric): Frame rate
789 • vcodec (string): Name of the video codec in use
791 • container (string): Name of the container format
793 • filesize (numeric): The number of bytes, if known in advance
795 • filesize_approx (numeric): An estimate for the number of bytes
797 • protocol (string): The protocol that will be used for the actual
799 download
800 • extractor (string): Name of the extractor
802 • extractor_key (string): Key name of the extractor
804 • epoch (numeric): Unix epoch when creating the file
806 • autonumber (numeric): Number that will be increased with each down‐
808 load, starting at --autonumber-start
809 • playlist (string): Name or id of the playlist that contains the video
811 • playlist_index (numeric): Index of the video in the playlist padded
813 with leading zeros according to the total length of the playlist
814 • playlist_id (string): Playlist identifier
816 • playlist_title (string): Playlist title
818 • playlist_uploader (string): Full name of the playlist uploader
820 • playlist_uploader_id (string): Nickname or id of the playlist upload‐
822 er
823 Available for the video that belongs to some logical chapter or sec‐
825 tion:
826 • chapter (string): Name or title of the chapter the video belongs to
828 • chapter_number (numeric): Number of the chapter the video belongs to
830 • chapter_id (string): Id of the chapter the video belongs to
832 Available for the video that is an episode of some series or programme:
834 • series (string): Title of the series or programme the video episode
836 belongs to
837 • season (string): Title of the season the video episode belongs to
839 • season_number (numeric): Number of the season the video episode be‐
841 longs to
842 • season_id (string): Id of the season the video episode belongs to
844 • episode (string): Title of the video episode
846 • episode_number (numeric): Number of the video episode within a season
848 • episode_id (string): Id of the video episode
850 Available for the media that is a track or a part of a music album:
852 • track (string): Title of the track
854 • track_number (numeric): Number of the track within an album or a disc
856 • track_id (string): Id of the track
858 • artist (string): Artist(s) of the track
860 • genre (string): Genre(s) of the track
862 • album (string): Title of the album the track belongs to
864 • album_type (string): Type of the album
866 • album_artist (string): List of all artists appeared on the album
868 • disc_number (numeric): Number of the disc or other physical medium
870 the track belongs to
871 • release_year (numeric): Year (YYYY) when the album was released
873 Each aforementioned sequence when referenced in an output template will
875 be replaced by the actual value corresponding to the sequence name.
876 Note that some of the sequences are not guaranteed to be present since
877 they depend on the metadata obtained by a particular extractor. Such
878 sequences will be replaced with placeholder value provided with --out‐
879 put-na-placeholder (NA by default).
880 For example for -o %(title)s-%(id)s.%(ext)s and an mp4 video with title
882 youtube-dl test video and id BaW_jenozKcj, this will result in a
883 youtube-dl test video-BaW_jenozKcj.mp4 file created in the current di‐
884 rectory.
885 For numeric sequences you can use numeric related formatting, for exam‐
887 ple, %(view_count)05d will result in a string with view count padded
888 with zeros up to 5 characters, like in 00042.
889 Output templates can also contain arbitrary hierarchical path, e.g. -o
891 '%(playlist)s/%(playlist_index)s - %(title)s.%(ext)s' which will result
892 in downloading each video in a directory corresponding to this path
893 template. Any missing directory will be automatically created for you.
894 To use percent literals in an output template use %%. To output to
896 stdout use -o -.
897 The current default template is %(title)s-%(id)s.%(ext)s.
899 In some cases, you don’t want special characters such as 中, spaces, or
901 &, such as when transferring the downloaded filename to a Windows sys‐
902 tem or the filename through an 8bit-unsafe channel. In these cases,
903 add the --restrict-filenames flag to get a shorter title.
904 Output template and Windows batch files
906 If you are using an output template inside a Windows batch file then
907 you must escape plain percent characters (%) by doubling, so that -o
908 "%(title)s-%(id)s.%(ext)s" should become -o "%%(ti‐
909 tle)s-%%(id)s.%%(ext)s". However you should not touch %’s that are not
910 plain characters, e.g. environment variables for expansion should stay
911 intact: -o "C:\%HOMEPATH%\Desktop\%%(title)s.%%(ext)s".
912 Output template examples
914 Note that on Windows you may need to use double quotes instead of sin‐
915 gle.
916 $ youtube-dl --get-filename -o '%(title)s.%(ext)s' BaW_jenozKc
918 youtube-dl test video ''_ä↭𝕐.mp4 # All kinds of weird characters
919 $ youtube-dl --get-filename -o '%(title)s.%(ext)s' BaW_jenozKc --restrict-filenames
921 youtube-dl_test_video_.mp4 # A simple file name
922 # Download YouTube playlist videos in separate directory indexed by video order in a playlist
924 $ youtube-dl -o '%(playlist)s/%(playlist_index)s - %(title)s.%(ext)s' https://www.youtube.com/playlist?list=PLwiyx1dc3P2JR9N8gQaQN_BCvlSlap7re
925 # Download all playlists of YouTube channel/user keeping each playlist in separate directory:
927 $ youtube-dl -o '%(uploader)s/%(playlist)s/%(playlist_index)s - %(title)s.%(ext)s' https://www.youtube.com/user/TheLinuxFoundation/playlists
928 # Download Udemy course keeping each chapter in separate directory under MyVideos directory in your home
930 $ youtube-dl -u user -p password -o '~/MyVideos/%(playlist)s/%(chapter_number)s - %(chapter)s/%(title)s.%(ext)s' https://www.udemy.com/java-tutorial/
931 # Download entire series season keeping each series and each season in separate directory under C:/MyVideos
933 $ youtube-dl -o "C:/MyVideos/%(series)s/%(season_number)s - %(season)s/%(episode_number)s - %(episode)s.%(ext)s" https://videomore.ru/kino_v_detalayah/5_sezon/367617
934 # Stream the video being downloaded to stdout
936 $ youtube-dl -o - BaW_jenozKc
937
939 By default youtube-dl tries to download the best available quality,
940 i.e. if you want the best quality you don’t need to pass any special
941 options, youtube-dl will guess it for you by default.
942 But sometimes you may want to download in a different format, for exam‐
944 ple when you are on a slow or intermittent connection. The key mecha‐
945 nism for achieving this is so-called format selection based on which
946 you can explicitly specify desired format, select formats based on some
947 criterion or criteria, setup precedence and much more.
948 The general syntax for format selection is --format FORMAT or shorter
950 -f FORMAT where FORMAT is a selector expression, i.e. an expression
951 that describes format or formats you would like to download.
952 tl;dr: navigate me to examples.
954 The simplest case is requesting a specific format, for example with -f
956 22 you can download the format with format code equal to 22. You can
957 get the list of available format codes for particular video using
958 --list-formats or -F. Note that these format codes are extractor spe‐
959 cific.
960 You can also use a file extension (currently 3gp, aac, flv, m4a, mp3,
962 mp4, ogg, wav, webm are supported) to download the best quality format
963 of a particular file extension served as a single file, e.g. -f webm
964 will download the best quality format with the webm extension served as
965 a single file.
966 You can also use special names to select particular edge case formats:
968 • best: Select the best quality format represented by a single file
970 with video and audio.
971 • worst: Select the worst quality format represented by a single file
973 with video and audio.
974 • bestvideo: Select the best quality video-only format (e.g. DASH
976 video). May not be available.
977 • worstvideo: Select the worst quality video-only format. May not be
979 available.
980 • bestaudio: Select the best quality audio only-format. May not be
982 available.
983 • worstaudio: Select the worst quality audio only-format. May not be
985 available.
986 For example, to download the worst quality video-only format you can
988 use -f worstvideo.
989 If you want to download multiple videos and they don’t have the same
991 formats available, you can specify the order of preference using slash‐
992 es. Note that slash is left-associative, i.e. formats on the left hand
993 side are preferred, for example -f 22/17/18 will download format 22 if
994 it’s available, otherwise it will download format 17 if it’s available,
995 otherwise it will download format 18 if it’s available, otherwise it
996 will complain that no suitable formats are available for download.
997 If you want to download several formats of the same video use a comma
999 as a separator, e.g. -f 22,17,18 will download all these three formats,
1000 of course if they are available. Or a more sophisticated example com‐
1001 bined with the precedence feature: -f
1002 136/137/mp4/bestvideo,140/m4a/bestaudio.
1003 You can also filter the video formats by putting a condition in brack‐
1005 ets, as in -f "best[height=720]" (or -f "[filesize>10M]").
1006 The following numeric meta fields can be used with comparisons <, <=,
1008 >, >=, = (equals), != (not equals):
1009 • filesize: The number of bytes, if known in advance
1011 • width: Width of the video, if known
1013 • height: Height of the video, if known
1015 • tbr: Average bitrate of audio and video in KBit/s
1017 • abr: Average audio bitrate in KBit/s
1019 • vbr: Average video bitrate in KBit/s
1021 • asr: Audio sampling rate in Hertz
1023 • fps: Frame rate
1025 Also filtering work for comparisons = (equals), ^= (starts with), $=
1027 (ends with), *= (contains) and following string meta fields:
1028 • ext: File extension
1030 • acodec: Name of the audio codec in use
1032 • vcodec: Name of the video codec in use
1034 • container: Name of the container format
1036 • protocol: The protocol that will be used for the actual download,
1038 lower-case (http, https, rtsp, rtmp, rtmpe, mms, f4m, ism,
1039 http_dash_segments, m3u8, or m3u8_native)
1040 • format_id: A short description of the format
1042 • language: Language code
1044 Any string comparison may be prefixed with negation ! in order to pro‐
1046 duce an opposite comparison, e.g. !*= (does not contain).
1047 Note that none of the aforementioned meta fields are guaranteed to be
1049 present since this solely depends on the metadata obtained by particu‐
1050 lar extractor, i.e. the metadata offered by the video hoster.
1051 Formats for which the value is not known are excluded unless you put a
1053 question mark (?) after the operator. You can combine format filters,
1054 so -f "[height <=? 720][tbr>500]" selects up to 720p videos (or videos
1055 where the height is not known) with a bitrate of at least 500 KBit/s.
1056 You can merge the video and audio of two formats into a single file us‐
1058 ing -f <video-format>+<audio-format> (requires ffmpeg or avconv in‐
1059 stalled), for example -f bestvideo+bestaudio will download the best
1060 video-only format, the best audio-only format and mux them together
1061 with ffmpeg/avconv.
1062 Format selectors can also be grouped using parentheses, for example if
1064 you want to download the best mp4 and webm formats with a height lower
1065 than 480 you can use -f '(mp4,webm)[height<480]'.
1066 Since the end of April 2015 and version 2015.04.26, youtube-dl uses -f
1068 bestvideo+bestaudio/best as the default format selection (see #5447
1069 (https://github.com/ytdl-org/youtube-dl/issues/5447), #5456
1070 (https://github.com/ytdl-org/youtube-dl/issues/5456)). If ffmpeg or
1071 avconv are installed this results in downloading bestvideo and bestau‐
1072 dio separately and muxing them together into a single file giving the
1073 best overall quality available. Otherwise it falls back to best and
1074 results in downloading the best available quality served as a single
1075 file. best is also needed for videos that don’t come from YouTube be‐
1076 cause they don’t provide the audio and video in two different files.
1077 If you want to only download some DASH formats (for example if you are
1078 not interested in getting videos with a resolution higher than 1080p),
1079 you can add -f bestvideo[height<=?1080]+bestaudio/best to your configu‐
1080 ration file. Note that if you use youtube-dl to stream to stdout (and
1081 most likely to pipe it to your media player then), i.e. you explicitly
1082 specify output template as -o -, youtube-dl still uses -f best format
1083 selection in order to start content delivery immediately to your player
1084 and not to wait until bestvideo and bestaudio are downloaded and muxed.
1085 If you want to preserve the old format selection behavior (prior to
1087 youtube-dl 2015.04.26), i.e. you want to download the best available
1088 quality media served as a single file, you should explicitly specify
1089 your choice with -f best. You may want to add it to the configuration
1090 file in order not to type it every time you run youtube-dl.
1091 Format selection examples
1093 Note that on Windows you may need to use double quotes instead of sin‐
1094 gle.
1095 # Download best mp4 format available or any other best if no mp4 available
1097 $ youtube-dl -f 'bestvideo[ext=mp4]+bestaudio[ext=m4a]/best[ext=mp4]/best'
1098 # Download best format available but no better than 480p
1100 $ youtube-dl -f 'bestvideo[height<=480]+bestaudio/best[height<=480]'
1101 # Download best video only format but no bigger than 50 MB
1103 $ youtube-dl -f 'best[filesize<50M]'
1104 # Download best format available via direct link over HTTP/HTTPS protocol
1106 $ youtube-dl -f '(bestvideo+bestaudio/best)[protocol^=http]'
1107 # Download the best video format and the best audio format without merging them
1109 $ youtube-dl -f 'bestvideo,bestaudio' -o '%(title)s.f%(format_id)s.%(ext)s'
1110 Note that in the last example, an output template is recommended as
1112 bestvideo and bestaudio may have the same file name.
1113
1115 Videos can be filtered by their upload date using the options --date,
1116 --datebefore or --dateafter. They accept dates in two formats:
1117 • Absolute dates: Dates in the format YYYYMMDD.
1119 • Relative dates: Dates in the format (now|to‐
1121 day)[+-][0-9](day|week|month|year)(s)?
1122 Examples:
1124 # Download only the videos uploaded in the last 6 months
1126 $ youtube-dl --dateafter now-6months
1127 # Download only the videos uploaded on January 1, 1970
1129 $ youtube-dl --date 19700101
1130 $ # Download only the videos uploaded in the 200x decade
1132 $ youtube-dl --dateafter 20000101 --datebefore 20091231
1133
1135 How do I update youtube-dl?
1136 If you’ve followed our manual installation instructions (https://ytdl-
1137 org.github.io/youtube-dl/download.html), you can simply run youtube-dl
1138 -U (or, on Linux, sudo youtube-dl -U).
1139 If you have used pip, a simple sudo pip install -U youtube-dl is suffi‐
1141 cient to update.
1142 If you have installed youtube-dl using a package manager like apt-get
1144 or yum, use the standard system update mechanism to update. Note that
1145 distribution packages are often outdated. As a rule of thumb, youtube-
1146 dl releases at least once a month, and often weekly or even daily.
1147 Simply go to https://yt-dl.org to find out the current version. Unfor‐
1148 tunately, there is nothing we youtube-dl developers can do if your dis‐
1149 tribution serves a really outdated version. You can (and should) com‐
1150 plain to your distribution in their bugtracker or support forum.
1151 As a last resort, you can also uninstall the version installed by your
1153 package manager and follow our manual installation instructions. For
1154 that, remove the distribution’s package, with a line like
1155 sudo apt-get remove -y youtube-dl
1157 Afterwards, simply follow our manual installation instructions
1159 (https://ytdl-org.github.io/youtube-dl/download.html):
1160 sudo wget https://yt-dl.org/downloads/latest/youtube-dl -O /usr/local/bin/youtube-dl
1162 sudo chmod a+rx /usr/local/bin/youtube-dl
1163 hash -r
1164 Again, from then on you’ll be able to update with sudo youtube-dl -U.
1166 youtube-dl is extremely slow to start on Windows
1168 Add a file exclusion for youtube-dl.exe in Windows Defender settings.
1169 I’m getting an error Unable to extract OpenGraph title on YouTube playlists
1171 YouTube changed their playlist format in March 2014 and later on, so
1172 you’ll need at least youtube-dl 2014.07.25 to download all YouTube
1173 videos.
1174 If you have installed youtube-dl with a package manager, pip, setup.py
1176 or a tarball, please use that to update. Note that Ubuntu packages do
1177 not seem to get updated anymore. Since we are not affiliated with
1178 Ubuntu, there is little we can do. Feel free to report bugs
1179 (https://bugs.launchpad.net/ubuntu/+source/youtube-dl/+filebug) to the
1180 Ubuntu packaging people (mailto:ubuntu-motu@lists.ubuntu.com?sub‐
1181 ject=outdated%20version%20of%20youtube-dl) - all they have to do is up‐
1182 date the package to a somewhat recent version. See above for a way to
1183 update.
1184 I’m getting an error when trying to use output template: error: using out‐
1186 put template conflicts with using title, video ID or auto number
1187 Make sure you are not using -o with any of these options -t, --title,
1188 --id, -A or --auto-number set in command line or in a configuration
1189 file. Remove the latter if any.
1190 Do I always have to pass -citw?
1192 By default, youtube-dl intends to have the best options (incidentally,
1193 if you have a convincing case that these should be different, please
1194 file an issue where you explain that (https://yt-dl.org/bug)). There‐
1195 fore, it is unnecessary and sometimes harmful to copy long option
1196 strings from webpages. In particular, the only option out of -citw
1197 that is regularly useful is -i.
1198 Can you please put the -b option back?
1200 Most people asking this question are not aware that youtube-dl now de‐
1201 faults to downloading the highest available quality as reported by
1202 YouTube, which will be 1080p or 720p in some cases, so you no longer
1203 need the -b option. For some specific videos, maybe YouTube does not
1204 report them to be available in a specific high quality format you’re
1205 interested in. In that case, simply request it with the -f option and
1206 youtube-dl will try to download it.
1207 I get HTTP error 402 when trying to download a video. What’s this?
1209 Apparently YouTube requires you to pass a CAPTCHA test if you download
1210 too much. We’re considering to provide a way to let you solve the
1211 CAPTCHA (https://github.com/ytdl-org/youtube-dl/issues/154), but at the
1212 moment, your best course of action is pointing a web browser to the
1213 youtube URL, solving the CAPTCHA, and restart youtube-dl.
1214 Do I need any other programs?
1216 youtube-dl works fine on its own on most sites. However, if you want
1217 to convert video/audio, you’ll need avconv (https://libav.org/) or ffm‐
1218 peg (https://www.ffmpeg.org/). On some sites - most notably YouTube -
1219 videos can be retrieved in a higher quality format without sound.
1220 youtube-dl will detect whether avconv/ffmpeg is present and automati‐
1221 cally pick the best option.
1222 Videos or video formats streamed via RTMP protocol can only be down‐
1224 loaded when rtmpdump (https://rtmpdump.mplayerhq.hu/) is installed.
1225 Downloading MMS and RTSP videos requires either mplayer (https://mplay‐
1226 erhq.hu/) or mpv (https://mpv.io/) to be installed.
1227 I have downloaded a video but how can I play it?
1229 Once the video is fully downloaded, use any video player, such as mpv
1230 (https://mpv.io/), vlc (https://www.videolan.org/) or mplayer
1231 (https://www.mplayerhq.hu/).
1232 I extracted a video URL with -g, but it does not play on another machine /
1234 in my web browser.
1235 It depends a lot on the service. In many cases, requests for the video
1236 (to download/play it) must come from the same IP address and with the
1237 same cookies and/or HTTP headers. Use the --cookies option to write
1238 the required cookies into a file, and advise your downloader to read
1239 cookies from that file. Some sites also require a common user agent to
1240 be used, use --dump-user-agent to see the one in use by youtube-dl.
1241 You can also get necessary cookies and HTTP headers from JSON output
1242 obtained with --dump-json.
1243 It may be beneficial to use IPv6; in some cases, the restrictions are
1245 only applied to IPv4. Some services (sometimes only for a subset of
1246 videos) do not restrict the video URL by IP address, cookie, or user-
1247 agent, but these are the exception rather than the rule.
1248 Please bear in mind that some URL protocols are not supported by
1250 browsers out of the box, including RTMP. If you are using -g, your own
1251 downloader must support these as well.
1252 If you want to play the video on a machine that is not running youtube-
1254 dl, you can relay the video content from the machine that runs youtube-
1255 dl. You can use -o - to let youtube-dl stream a video to stdout, or
1256 simply allow the player to download the files written by youtube-dl in
1257 turn.
1258 ERROR: no fmt_url_map or conn information found in video info
1260 YouTube has switched to a new video info format in July 2011 which is
1261 not supported by old versions of youtube-dl. See above for how to up‐
1262 date youtube-dl.
1263 ERROR: unable to download video
1265 YouTube requires an additional signature since September 2012 which is
1266 not supported by old versions of youtube-dl. See above for how to up‐
1267 date youtube-dl.
1268 Video URL contains an ampersand and I’m getting some strange output [1]
1270 2839 or 'v' is not recognized as an internal or external command
1271 That’s actually the output from your shell. Since ampersand is one of
1272 the special shell characters it’s interpreted by the shell preventing
1273 you from passing the whole URL to youtube-dl. To disable your shell
1274 from interpreting the ampersands (or any other special characters) you
1275 have to either put the whole URL in quotes or escape them with a back‐
1276 slash (which approach will work depends on your shell).
1277 For example if your URL is
1279 https://www.youtube.com/watch?t=4&v=BaW_jenozKc you should end up with
1280 following command:
1281 youtube-dl 'https://www.youtube.com/watch?t=4&v=BaW_jenozKc'
1283 or
1285 youtube-dl https://www.youtube.com/watch?t=4\&v=BaW_jenozKc
1287 For Windows you have to use the double quotes:
1289 youtube-dl "https://www.youtube.com/watch?t=4&v=BaW_jenozKc"
1291 ExtractorError: Could not find JS function u’OF’
1293 In February 2015, the new YouTube player contained a character sequence
1294 in a string that was misinterpreted by old versions of youtube-dl. See
1295 above for how to update youtube-dl.
1296 HTTP Error 429: Too Many Requests or 402: Payment Required
1298 These two error codes indicate that the service is blocking your IP ad‐
1299 dress because of overuse. Usually this is a soft block meaning that
1300 you can gain access again after solving CAPTCHA. Just open a browser
1301 and solve a CAPTCHA the service suggests you and after that pass cook‐
1302 ies to youtube-dl. Note that if your machine has multiple external IPs
1303 then you should also pass exactly the same IP you’ve used for solving
1304 CAPTCHA with --source-address. Also you may need to pass a User-Agent
1305 HTTP header of your browser with --user-agent.
1306 If this is not the case (no CAPTCHA suggested to solve by the service)
1308 then you can contact the service and ask them to unblock your IP ad‐
1309 dress, or - if you have acquired a whitelisted IP address already - use
1310 the --proxy or --source-address options to select another IP address.
1311 SyntaxError: Non-ASCII character
1313 The error
1314 File "youtube-dl", line 2
1316 SyntaxError: Non-ASCII character '\x93' ...
1317 means you’re using an outdated version of Python. Please update to
1319 Python 2.6 or 2.7.
1320 What is this binary file? Where has the code gone?
1322 Since June 2012 (#342 (https://github.com/ytdl-org/youtube-dl/is‐
1323 sues/342)) youtube-dl is packed as an executable zipfile, simply unzip
1324 it (might need renaming to youtube-dl.zip first on some systems) or
1325 clone the git repository, as laid out above. If you modify the code,
1326 you can run it by executing the __main__.py file. To recompile the ex‐
1327 ecutable, run make youtube-dl.
1328 The exe throws an error due to missing MSVCR100.dll
1330 To run the exe you need to install first the Microsoft Visual C++ 2010
1331 Service Pack 1 Redistributable Package (x86) (https://download.micro‐
1332 soft.com/download/1/6/5/165255E7-1014-4D0A-B094-B6A430A6BFFC/vcre‐
1333 dist_x86.exe).
1334 On Windows, how should I set up ffmpeg and youtube-dl? Where should I put
1336 the exe files?
1337 If you put youtube-dl and ffmpeg in the same directory that you’re run‐
1338 ning the command from, it will work, but that’s rather cumbersome.
1339 To make a different directory work - either for ffmpeg, or for youtube-
1341 dl, or for both - simply create the directory (say, C:\bin, or
1342 C:\Users\<User name>\bin), put all the executables directly in there,
1343 and then set your PATH environment variable (https://www.ja‐
1344 va.com/en/download/help/path.xml) to include that directory.
1345 From then on, after restarting your shell, you will be able to access
1347 both youtube-dl and ffmpeg (and youtube-dl will be able to find ffmpeg)
1348 by simply typing youtube-dl or ffmpeg, no matter what directory you’re
1349 in.
1350 How do I put downloads into a specific folder?
1352 Use the -o to specify an output template, for example -o "/home/us‐
1353 er/videos/%(title)s-%(id)s.%(ext)s". If you want this for all of your
1354 downloads, put the option into your configuration file.
1355 How do I download a video starting with a -?
1357 Either prepend https://www.youtube.com/watch?v= or separate the ID from
1358 the options with --:
1359 youtube-dl -- -wNyEUrxzFU
1361 youtube-dl "https://www.youtube.com/watch?v=-wNyEUrxzFU"
1362 How do I pass cookies to youtube-dl?
1364 Use the --cookies option, for example --cookies /path/to/cook‐
1365 ies/file.txt.
1366 In order to extract cookies from browser use any conforming browser ex‐
1368 tension for exporting cookies. For example, Get cookies.txt LOCALLY
1369 (https://chrome.google.com/webstore/detail/get-cookiestxt-local‐
1370 ly/cclelndahbckbenkjhflpdbgdldlbecc) (for Chrome) or cookies.txt
1371 (https://addons.mozilla.org/en-US/firefox/addon/cookies-txt/) (for
1372 Firefox).
1373 Note that the cookies file must be in Mozilla/Netscape format and the
1375 first line of the cookies file must be either # HTTP Cookie File or #
1376 Netscape HTTP Cookie File. Make sure you have correct newline format
1377 (https://en.wikipedia.org/wiki/Newline) in the cookies file and convert
1378 newlines if necessary to correspond with your OS, namely CRLF (\r\n)
1379 for Windows and LF (\n) for Unix and Unix-like systems (Linux, macOS,
1380 etc.). HTTP Error 400: Bad Request when using --cookies is a good sign
1381 of invalid newline format.
1382 Passing cookies to youtube-dl is a good way to workaround login when a
1384 particular extractor does not implement it explicitly. Another use
1385 case is working around CAPTCHA (https://en.wikipedia.org/wiki/CAPTCHA)
1386 some websites require you to solve in particular cases in order to get
1387 access (e.g. YouTube, CloudFlare).
1388 How do I stream directly to media player?
1390 You will first need to tell youtube-dl to stream media to stdout with
1391 -o -, and also tell your media player to read from stdin (it must be
1392 capable of this for streaming) and then pipe former to latter. For ex‐
1393 ample, streaming to vlc (https://www.videolan.org/) can be achieved
1394 with:
1395 youtube-dl -o - "https://www.youtube.com/watch?v=BaW_jenozKcj" | vlc -
1397 How do I download only new videos from a playlist?
1399 Use download-archive feature. With this feature you should initially
1400 download the complete playlist with --download-archive /path/to/down‐
1401 load/archive/file.txt that will record identifiers of all the videos in
1402 a special file. Each subsequent run with the same --download-archive
1403 will download only new videos and skip all videos that have been down‐
1404 loaded before. Note that only successful downloads are recorded in the
1405 file.
1406 For example, at first,
1408 youtube-dl --download-archive archive.txt "https://www.youtube.com/playlist?list=PLwiyx1dc3P2JR9N8gQaQN_BCvlSlap7re"
1410 will download the complete PLwiyx1dc3P2JR9N8gQaQN_BCvlSlap7re playlist
1412 and create a file archive.txt. Each subsequent run will only download
1413 new videos if any:
1414 youtube-dl --download-archive archive.txt "https://www.youtube.com/playlist?list=PLwiyx1dc3P2JR9N8gQaQN_BCvlSlap7re"
1416 Should I add --hls-prefer-native into my config?
1418 When youtube-dl detects an HLS video, it can download it either with
1419 the built-in downloader or ffmpeg. Since many HLS streams are slightly
1420 invalid and ffmpeg/youtube-dl each handle some invalid cases better
1421 than the other, there is an option to switch the downloader if needed.
1422 When youtube-dl knows that one particular downloader works better for a
1424 given website, that downloader will be picked. Otherwise, youtube-dl
1425 will pick the best downloader for general compatibility, which at the
1426 moment happens to be ffmpeg. This choice may change in future versions
1427 of youtube-dl, with improvements of the built-in downloader and/or ffm‐
1428 peg.
1429 In particular, the generic extractor (used when your website is not in
1431 the list of supported sites by youtube-dl (https://ytdl-
1432 org.github.io/youtube-dl/supportedsites.html) cannot mandate one spe‐
1433 cific downloader.
1434 If you put either --hls-prefer-native or --hls-prefer-ffmpeg into your
1436 configuration, a different subset of videos will fail to download cor‐
1437 rectly. Instead, it is much better to file an issue (https://yt-
1438 dl.org/bug) or a pull request which details why the native or the ffm‐
1439 peg HLS downloader is a better choice for your use case.
1440 Can you add support for this anime video site, or site which shows current
1442 movies for free?
1443 As a matter of policy (as well as legality), youtube-dl does not in‐
1444 clude support for services that specialize in infringing copyright. As
1445 a rule of thumb, if you cannot easily find a video that the service is
1446 quite obviously allowed to distribute (i.e. that has been uploaded by
1447 the creator, the creator’s distributor, or is published under a free
1448 license), the service is probably unfit for inclusion to youtube-dl.
1449 A note on the service that they don’t host the infringing content, but
1451 just link to those who do, is evidence that the service should not be
1452 included into youtube-dl. The same goes for any DMCA note when the
1453 whole front page of the service is filled with videos they are not al‐
1454 lowed to distribute. A “fair use” note is equally unconvincing if the
1455 service shows copyright-protected videos in full without authorization.
1456 Support requests for services that do purchase the rights to distribute
1458 their content are perfectly fine though. If in doubt, you can simply
1459 include a source that mentions the legitimate purchase of content.
1460 How can I speed up work on my issue?
1462 (Also known as: Help, my important issue not being solved!) The
1463 youtube-dl core developer team is quite small. While we do our best to
1464 solve as many issues as possible, sometimes that can take quite a
1465 while. To speed up your issue, here’s what you can do:
1466 First of all, please do report the issue at our issue tracker
1468 (https://yt-dl.org/bugs). That allows us to coordinate all efforts by
1469 users and developers, and serves as a unified point. Unfortunately,
1470 the youtube-dl project has grown too large to use personal email as an
1471 effective communication channel.
1472 Please read the bug reporting instructions below. A lot of bugs lack
1474 all the necessary information. If you can, offer proxy, VPN, or shell
1475 access to the youtube-dl developers. If you are able to, test the is‐
1476 sue from multiple computers in multiple countries to exclude local cen‐
1477 sorship or misconfiguration issues.
1478 If nobody is interested in solving your issue, you are welcome to take
1480 matters into your own hands and submit a pull request (or coerce/pay
1481 somebody else to do so).
1482 Feel free to bump the issue from time to time by writing a small com‐
1484 ment (“Issue is still present in youtube-dl version ...from France, but
1485 fixed from Belgium”), but please not more than once a month. Please do
1486 not declare your issue as important or urgent.
1487 How can I detect whether a given URL is supported by youtube-dl?
1489 For one, have a look at the list of supported sites. Note that it can
1490 sometimes happen that the site changes its URL scheme (say, from
1491 https://example.com/video/1234567 to https://example.com/v/1234567 )
1492 and youtube-dl reports an URL of a service in that list as unsupported.
1493 In that case, simply report a bug.
1494 It is not possible to detect whether a URL is supported or not. That’s
1496 because youtube-dl contains a generic extractor which matches all URLs.
1497 You may be tempted to disable, exclude, or remove the generic extrac‐
1498 tor, but the generic extractor not only allows users to extract videos
1499 from lots of websites that embed a video from another service, but may
1500 also be used to extract video from a service that it’s hosting itself.
1501 Therefore, we neither recommend nor support disabling, excluding, or
1502 removing the generic extractor.
1503 If you want to find out whether a given URL is supported, simply call
1505 youtube-dl with it. If you get no videos back, chances are the URL is
1506 either not referring to a video or unsupported. You can find out which
1507 by examining the output (if you run youtube-dl on the console) or
1508 catching an UnsupportedError exception if you run it from a Python pro‐
1509 gram.
1510
1512 Before we had the issue template, despite our extensive bug reporting
1513 instructions, about 80% of the issue reports we got were useless, for
1514 instance because people used ancient versions hundreds of releases old,
1515 because of simple syntactic errors (not in youtube-dl but in general
1516 shell usage), because the problem was already reported multiple times
1517 before, because people did not actually read an error message, even if
1518 it said “please install ffmpeg”, because people did not mention the URL
1519 they were trying to download and many more simple, easy-to-avoid prob‐
1520 lems, many of whom were totally unrelated to youtube-dl.
1521 youtube-dl is an open-source project manned by too few volunteers, so
1523 we’d rather spend time fixing bugs where we are certain none of those
1524 simple problems apply, and where we can be reasonably confident to be
1525 able to reproduce the issue without asking the reporter repeatedly. As
1526 such, the output of youtube-dl -v YOUR_URL_HERE is really all that’s
1527 required to file an issue. The issue template also guides you through
1528 some basic steps you can do, such as checking that your version of
1529 youtube-dl is current.
1530
1532 Most users do not need to build youtube-dl and can download the builds
1533 (https://ytdl-org.github.io/youtube-dl/download.html) or get them from
1534 their distribution.
1535 To run youtube-dl as a developer, you don’t need to build anything ei‐
1537 ther. Simply execute
1538 python -m youtube_dl
1540 To run the test, simply invoke your favorite test runner, or execute a
1542 test file directly; any of the following work:
1543 python -m unittest discover
1545 python test/test_download.py
1546 nosetests
1547 For Python versions 3.6 and later, you can use pynose
1549 (https://pypi.org/project/pynose/) to implement nosetests. The origi‐
1550 nal nose (https://pypi.org/project/nose/) has not been upgraded for
1551 3.10 and later.
1552 See item 6 of new extractor tutorial for how to run extractor specific
1554 test cases.
1555 If you want to create a build of youtube-dl yourself, you’ll need
1557 • python
1559 • make (only GNU make is supported)
1561 • pandoc
1563 • zip
1565 • nosetests
1567 Adding support for a new site
1569 If you want to add support for a new site, first of all make sure this
1570 site is not dedicated to copyright infringement. youtube-dl does not
1571 support such sites thus pull requests adding support for them will be
1572 rejected.
1573 After you have ensured this site is distributing its content legally,
1575 you can follow this quick list (assuming your service is called yourex‐
1576 tractor):
1577 1. Fork this repository (https://github.com/ytdl-org/youtube-dl/fork)
1579 2. Check out the source code with:
1581 git clone git@github.com:YOUR_GITHUB_USERNAME/youtube-dl.git
1583 3. Start a new git branch with
1585 cd youtube-dl
1587 git checkout -b yourextractor
1588 4. Start with this simple template and save it to youtube_dl/extrac‐
1590 tor/yourextractor.py:
1591 # coding: utf-8
1593 from __future__ import unicode_literals
1594 from .common import InfoExtractor
1596 class YourExtractorIE(InfoExtractor):
1599 _VALID_URL = r'https?://(?:www\.)?yourextractor\.com/watch/(?P<id>[0-9]+)'
1600 _TEST = {
1601 'url': 'https://yourextractor.com/watch/42',
1602 'md5': 'TODO: md5 sum of the first 10241 bytes of the video file (use --test)',
1603 'info_dict': {
1604 'id': '42',
1605 'ext': 'mp4',
1606 'title': 'Video title goes here',
1607 'thumbnail': r're:^https?://.*\.jpg$',
1608 # TODO more properties, either as:
1609 # * A value
1610 # * MD5 checksum; start the string with md5:
1611 # * A regular expression; start the string with re:
1612 # * Any Python type (for example int or float)
1613 }
1614 }
1615 def _real_extract(self, url):
1617 video_id = self._match_id(url)
1618 webpage = self._download_webpage(url, video_id)
1619 # TODO more code goes here, for example ...
1621 title = self._html_search_regex(r'<h1>(.+?)</h1>', webpage, 'title')
1622 return {
1624 'id': video_id,
1625 'title': title,
1626 'description': self._og_search_description(webpage),
1627 'uploader': self._search_regex(r'<div[^>]+id="uploader"[^>]*>([^<]+)<', webpage, 'uploader', fatal=False),
1628 # TODO more properties (see youtube_dl/extractor/common.py)
1629 }
1630 5. Add an import in youtube_dl/extractor/extractors.py
1632 (https://github.com/ytdl-org/youtube-dl/blob/master/youtube_dl/ex‐
1633 tractor/extractors.py).
1634 6. Run python test/test_download.py TestDownload.test_YourExtractor.
1636 This should fail at first, but you can continually re-run it until
1637 you’re done. If you decide to add more than one test (actually,
1638 test case) then rename _TEST to _TESTS and make it into a list of
1639 dictionaries. The tests will then be named TestDown‐
1640 load.test_YourExtractor, TestDownload.test_YourExtractor_1, Test‐
1641 Download.test_YourExtractor_2, etc. Note:
1642 • the test names use the extractor class name without the trailing
1644 IE
1645 • tests with only_matching key in test’s dict are not counted.
1647 7. Have a look at youtube_dl/extractor/common.py
1649 (https://github.com/ytdl-org/youtube-dl/blob/master/youtube_dl/ex‐
1650 tractor/common.py) for possible helper methods and a detailed de‐
1651 scription of what your extractor should and may return
1652 (https://github.com/ytdl-org/youtube-
1653 dl/blob/7f41a598b3fba1bcab2817de64a08941200aa3c8/youtube_dl/extrac‐
1654 tor/common.py#L94-L303). Add tests and code for as many as you
1655 want.
1656 8. Make sure your code follows youtube-dl coding conventions and check
1658 the code with flake8 (https://flake8.pycqa.org/en/latest/in‐
1659 dex.html#quickstart):
1660 $ flake8 youtube_dl/extractor/yourextractor.py
1662 9. Make sure your code works under all Python
1664 (https://www.python.org/) versions claimed supported by youtube-dl,
1665 namely 2.6, 2.7, and 3.2+.
1666 10. When the tests pass, add (https://git-scm.com/docs/git-add) the new
1668 files and commit (https://git-scm.com/docs/git-commit) them and
1669 push (https://git-scm.com/docs/git-push) the result, like this:
1670 $ git add youtube_dl/extractor/extractors.py
1672 $ git add youtube_dl/extractor/yourextractor.py
1673 $ git commit -m '[yourextractor] Add new extractor'
1674 $ git push origin yourextractor
1675 11. Finally, create a pull request (https://help.github.com/arti‐
1677 cles/creating-a-pull-request). We’ll then review and merge it.
1678 In any case, thank you very much for your contributions!
1680 youtube-dl coding conventions
1682 This section introduces guidelines for writing idiomatic, robust and
1683 future-proof extractor code.
1684 Extractors are very fragile by nature since they depend on the layout
1686 of the source data provided by 3rd party media hosters out of your con‐
1687 trol and this layout tends to change. As an extractor implementer your
1688 task is not only to write code that will extract media links and meta‐
1689 data correctly but also to minimize dependency on the source’s layout
1690 and even to make the code foresee potential future changes and be ready
1691 for that. This is important because it will allow the extractor not to
1692 break on minor layout changes thus keeping old youtube-dl versions
1693 working. Even though this breakage issue is easily fixed by emitting a
1694 new version of youtube-dl with a fix incorporated, all the previous
1695 versions become broken in all repositories and distros’ packages that
1696 may not be so prompt in fetching the update from us. Needless to say,
1697 some non rolling release distros may never receive an update at all.
1698 Mandatory and optional metafields
1700 For extraction to work youtube-dl relies on metadata your extractor ex‐
1701 tracts and provides to youtube-dl expressed by an information dictio‐
1702 nary (https://github.com/ytdl-org/youtube-
1703 dl/blob/7f41a598b3fba1bcab2817de64a08941200aa3c8/youtube_dl/extrac‐
1704 tor/common.py#L94-L303) or simply info dict. Only the following meta
1705 fields in the info dict are considered mandatory for a successful ex‐
1706 traction process by youtube-dl:
1707 • id (media identifier)
1709 • title (media title)
1711 • url (media download URL) or formats
1713 In fact only the last option is technically mandatory (i.e. if you
1715 can’t figure out the download location of the media the extraction does
1716 not make any sense). But by convention youtube-dl also treats id and
1717 title as mandatory. Thus the aforementioned metafields are the criti‐
1718 cal data that the extraction does not make any sense without and if any
1719 of them fail to be extracted then the extractor is considered complete‐
1720 ly broken.
1721 Any field (https://github.com/ytdl-org/youtube-
1723 dl/blob/7f41a598b3fba1bcab2817de64a08941200aa3c8/youtube_dl/extrac‐
1724 tor/common.py#L188-L303) apart from the aforementioned ones are consid‐
1725 ered optional. That means that extraction should be tolerant to situa‐
1726 tions when sources for these fields can potentially be unavailable
1727 (even if they are always available at the moment) and future-proof in
1728 order not to break the extraction of general purpose mandatory fields.
1729 Example
1731 Say you have some source dictionary meta that you’ve fetched as JSON
1732 with HTTP request and it has a key summary:
1733 meta = self._download_json(url, video_id)
1735 Assume at this point meta’s layout is:
1737 {
1739 ...
1740 "summary": "some fancy summary text",
1741 ...
1742 }
1743 Assume you want to extract summary and put it into the resulting info
1745 dict as description. Since description is an optional meta field you
1746 should be ready that this key may be missing from the meta dict, so
1747 that you should extract it like:
1748 description = meta.get('summary') # correct
1750 and not like:
1752 description = meta['summary'] # incorrect
1754 The latter will break extraction process with KeyError if summary dis‐
1756 appears from meta at some later time but with the former approach ex‐
1757 traction will just go ahead with description set to None which is per‐
1758 fectly fine (remember None is equivalent to the absence of data).
1759 Similarly, you should pass fatal=False when extracting optional data
1761 from a webpage with _search_regex, _html_search_regex or similar meth‐
1762 ods, for instance:
1763 description = self._search_regex(
1765 r'<span[^>]+id="title"[^>]*>([^<]+)<',
1766 webpage, 'description', fatal=False)
1767 With fatal set to False if _search_regex fails to extract description
1769 it will emit a warning and continue extraction.
1770 You can also pass default=<some fallback value>, for example:
1772 description = self._search_regex(
1774 r'<span[^>]+id="title"[^>]*>([^<]+)<',
1775 webpage, 'description', default=None)
1776 On failure this code will silently continue the extraction with de‐
1778 scription set to None. That is useful for metafields that may or may
1779 not be present.
1780 Provide fallbacks
1782 When extracting metadata try to do so from multiple sources. For exam‐
1783 ple if title is present in several places, try extracting from at least
1784 some of them. This makes it more future-proof in case some of the
1785 sources become unavailable.
1786 Example
1788 Say meta from the previous example has a title and you are about to ex‐
1789 tract it. Since title is a mandatory meta field you should end up with
1790 something like:
1791 title = meta['title']
1793 If title disappears from meta in future due to some changes on the
1795 hoster’s side the extraction would fail since title is mandatory.
1796 That’s expected.
1797 Assume that you have some another source you can extract title from,
1799 for example og:title HTML meta of a webpage. In this case you can pro‐
1800 vide a fallback scenario:
1801 title = meta.get('title') or self._og_search_title(webpage)
1803 This code will try to extract from meta first and if it fails it will
1805 try extracting og:title from a webpage.
1806 Regular expressions
1808 Don’t capture groups you don’t use
1809 Capturing group must be an indication that it’s used somewhere in the
1810 code. Any group that is not used must be non capturing.
1811 Example
1813 Don’t capture id attribute name here since you can’t use it for any‐
1814 thing anyway.
1815 Correct:
1817 r'(?:id|ID)=(?P<id>\d+)'
1819 Incorrect:
1821 r'(id|ID)=(?P<id>\d+)'
1823 Make regular expressions relaxed and flexible
1825 When using regular expressions try to write them fuzzy, relaxed and
1826 flexible, skipping insignificant parts that are more likely to change,
1827 allowing both single and double quotes for quoted values and so on.
1828 Example
1830 Say you need to extract title from the following HTML code:
1831 <span style="position: absolute; left: 910px; width: 90px; float: right; z-index: 9999;" class="title">some fancy title</span>
1833 The code for that task should look similar to:
1835 title = self._search_regex(
1837 r'<span[^>]+class="title"[^>]*>([^<]+)', webpage, 'title')
1838 Or even better:
1840 title = self._search_regex(
1842 r'<span[^>]+class=(["\'])title\1[^>]*>(?P<title>[^<]+)',
1843 webpage, 'title', group='title')
1844 Note how you tolerate potential changes in the style attribute’s value
1846 or switch from using double quotes to single for class attribute:
1847 The code definitely should not look like:
1849 title = self._search_regex(
1851 r'<span style="position: absolute; left: 910px; width: 90px; float: right; z-index: 9999;" class="title">(.*?)</span>',
1852 webpage, 'title', group='title')
1853 Long lines policy
1855 There is a soft limit to keep lines of code under 80 characters long.
1856 This means it should be respected if possible and if it does not make
1857 readability and code maintenance worse.
1858 For example, you should never split long string literals like URLs or
1860 some other often copied entities over multiple lines to fit this limit:
1861 Correct:
1863 'https://www.youtube.com/watch?v=FqZTN594JQw&list=PLMYEtVRpaqY00V9W81Cwmzp6N6vZqfUKD4'
1865 Incorrect:
1867 'https://www.youtube.com/watch?v=FqZTN594JQw&list='
1869 'PLMYEtVRpaqY00V9W81Cwmzp6N6vZqfUKD4'
1870 Inline values
1872 Extracting variables is acceptable for reducing code duplication and
1873 improving readability of complex expressions. However, you should
1874 avoid extracting variables used only once and moving them to opposite
1875 parts of the extractor file, which makes reading the linear flow diffi‐
1876 cult.
1877 Example
1879 Correct:
1880 title = self._html_search_regex(r'<title>([^<]+)</title>', webpage, 'title')
1882 Incorrect:
1884 TITLE_RE = r'<title>([^<]+)</title>'
1886 # ...some lines of code...
1887 title = self._html_search_regex(TITLE_RE, webpage, 'title')
1888 Collapse fallbacks
1890 Multiple fallback values can quickly become unwieldy. Collapse multi‐
1891 ple fallback values into a single expression via a list of patterns.
1892 Example
1894 Good:
1895 description = self._html_search_meta(
1897 ['og:description', 'description', 'twitter:description'],
1898 webpage, 'description', default=None)
1899 Unwieldy:
1901 description = (
1903 self._og_search_description(webpage, default=None)
1904 or self._html_search_meta('description', webpage, default=None)
1905 or self._html_search_meta('twitter:description', webpage, default=None))
1906 Methods supporting list of patterns are: _search_regex,
1908 _html_search_regex, _og_search_property, _html_search_meta.
1909 Trailing parentheses
1911 Always move trailing parentheses after the last argument.
1912 Example
1914 Correct:
1915 lambda x: x['ResultSet']['Result'][0]['VideoUrlSet']['VideoUrl'],
1917 list)
1918 Incorrect:
1920 lambda x: x['ResultSet']['Result'][0]['VideoUrlSet']['VideoUrl'],
1922 list,
1923 )
1924 Use convenience conversion and parsing functions
1926 Wrap all extracted numeric data into safe functions from
1927 youtube_dl/utils.py (https://github.com/ytdl-org/youtube-dl/blob/mas‐
1928 ter/youtube_dl/utils.py): int_or_none, float_or_none. Use them for
1929 string to number conversions as well.
1930 Use url_or_none for safe URL processing.
1932 Use traverse_obj for safe metadata extraction from parsed JSON.
1934 Use unified_strdate for uniform upload_date or any YYYYMMDD meta field
1936 extraction, unified_timestamp for uniform timestamp extraction,
1937 parse_filesize for filesize extraction, parse_count for count meta
1938 fields extraction, parse_resolution, parse_duration for duration ex‐
1939 traction, parse_age_limit for age_limit extraction.
1940 Explore youtube_dl/utils.py (https://github.com/ytdl-org/youtube-
1942 dl/blob/master/youtube_dl/utils.py) for more useful convenience func‐
1943 tions.
1944 More examples
1946 Safely extract optional description from parsed JSON
1947 When processing complex JSON, as often returned by site API requests or
1948 stashed in web pages for “hydration”, you can use the traverse_obj()
1949 utility function to handle multiple fallback values and to ensure the
1950 expected type of metadata items. The function’s docstring defines how
1951 the function works: also review usage in the codebase for more exam‐
1952 ples.
1953 In this example, a text description, or None, is pulled from the .re‐
1955 sult.video[0].summary member of the parsed JSON response, if available.
1956 description = traverse_obj(response, ('result', 'video', 0, 'summary', T(compat_str)))
1958 T(...) is a shorthand for a set literal; if you hate people who still
1960 run Python 2.6, T(type_or_transformation) could be written as a set
1961 literal {type_or_transformation}.
1962 Some extractors use the older and less capable try_get() function in
1964 the same way.
1965 description = try_get(response, lambda x: x['result']['video'][0]['summary'], compat_str)
1967 Safely extract more optional metadata
1969 In this example, various optional metadata values are extracted from
1970 the .result.video[0] member of the parsed JSON response, which is ex‐
1971 pected to be a JS object, parsed into a dict, with no crash if that
1972 isn’t so, or if any of the target values are missing or invalid.
1973 video = traverse_obj(response, ('result', 'video', 0, T(dict))) or {}
1975 # formerly:
1976 # video = try_get(response, lambda x: x['result']['video'][0], dict) or {}
1977 description = video.get('summary')
1978 duration = float_or_none(video.get('durationMs'), scale=1000)
1979 view_count = int_or_none(video.get('views'))
1980 Safely extract nested lists
1982 Suppose you’ve extracted JSON like this into a Python data structure
1983 named media_json using, say, the _download_json() or _parse_json()
1984 methods of InfoExtractor:
1985 {
1987 "title": "Example video",
1988 "comment": "try extracting this",
1989 "media": [{
1990 "type": "bad",
1991 "size": 320,
1992 "url": "https://some.cdn.site/bad.mp4"
1993 }, {
1994 "type": "streaming",
1995 "url": "https://some.cdn.site/hls.m3u8"
1996 }, {
1997 "type": "super",
1998 "size": 1280,
1999 "url": "https://some.cdn.site/good.webm"
2000 }],
2001 "moreStuff": "more values",
2002 ...
2003 }
2004 Then extractor code like this can collect the various fields of the
2006 JSON:
2007 ...
2009 from ..utils import (
2010 determine_ext,
2011 int_or_none,
2012 T,
2013 traverse_obj,
2014 txt_or_none,
2015 url_or_none,
2016 )
2017 ...
2018 ...
2019 info_dict = {}
2020 # extract title and description if valid and not empty
2021 info_dict.update(traverse_obj(media_json, {
2022 'title': ('title', T(txt_or_none)),
2023 'description': ('comment', T(txt_or_none)),
2024 }))
2025 # extract any recognisable media formats
2027 fmts = []
2028 # traverse into "media" list, extract `dict`s with desired keys
2029 for fmt in traverse_obj(media_json, ('media', Ellipsis, {
2030 'format_id': ('type', T(txt_or_none)),
2031 'url': ('url', T(url_or_none)),
2032 'width': ('size', T(int_or_none)), })):
2033 # bad `fmt` values were `None` and removed
2034 if 'url' not in fmt:
2035 continue
2036 fmt_url = fmt['url'] # known to be valid URL
2037 ext = determine_ext(fmt_url)
2038 if ext == 'm3u8':
2039 fmts.extend(self._extract_m3u8_formats(fmt_url, video_id, 'mp4', fatal=False))
2040 else:
2041 fmt['ext'] = ext
2042 fmts.append(fmt)
2043 # sort, raise if no formats
2045 self._sort_formats(fmts)
2046 info_dict['formats'] = fmts
2048 ...
2049 The extractor raises an exception rather than random crashes if the
2051 JSON structure changes so that no formats are found.
2052
2054 youtube-dl makes the best effort to be a good command-line program, and
2055 thus should be callable from any programming language. If you en‐
2056 counter any problems parsing its output, feel free to create a report
2057 (https://github.com/ytdl-org/youtube-dl/issues/new).
2058 From a Python program, you can embed youtube-dl in a more powerful
2060 fashion, like this:
2061 from __future__ import unicode_literals
2063 import youtube_dl
2064 ydl_opts = {}
2066 with youtube_dl.YoutubeDL(ydl_opts) as ydl:
2067 ydl.download(['https://www.youtube.com/watch?v=BaW_jenozKc'])
2068 Most likely, you’ll want to use various options. For a list of options
2070 available, have a look at youtube_dl/YoutubeDL.py
2071 (https://github.com/ytdl-org/youtube-
2072 dl/blob/3e4cedf9e8cd3157df2457df7274d0c842421945/youtube_dl/YoutubeDL.py#L137-L312).
2073 For a start, if you want to intercept youtube-dl’s output, set a logger
2074 object.
2075 Here’s a more complete example of a program that outputs only errors
2077 (and a short message after the download is finished), and down‐
2078 loads/converts the video to an mp3 file:
2079 from __future__ import unicode_literals
2081 import youtube_dl
2082 class MyLogger(object):
2085 def debug(self, msg):
2086 pass
2087 def warning(self, msg):
2089 pass
2090 def error(self, msg):
2092 print(msg)
2093 def my_hook(d):
2096 if d['status'] == 'finished':
2097 print('Done downloading, now converting ...')
2098 ydl_opts = {
2101 'format': 'bestaudio/best',
2102 'postprocessors': [{
2103 'key': 'FFmpegExtractAudio',
2104 'preferredcodec': 'mp3',
2105 'preferredquality': '192',
2106 }],
2107 'logger': MyLogger(),
2108 'progress_hooks': [my_hook],
2109 }
2110 with youtube_dl.YoutubeDL(ydl_opts) as ydl:
2111 ydl.download(['https://www.youtube.com/watch?v=BaW_jenozKc'])
2112
2114 Bugs and suggestions should be reported in the issue tracker:
2115 <https://github.com/ytdl-org/youtube-dl/issues> (<https://yt-
2116 dl.org/bug> is an alias for this). Unless you were prompted to or
2117 there is another pertinent reason (e.g. GitHub fails to accept the bug
2118 report), please do not send bug reports via personal email. For dis‐
2119 cussions, join us in the IRC channel #youtube-dl (irc://chat.freen‐
2120 ode.net/#youtube-dl) on freenode (webchat (https://webchat.freen‐
2121 ode.net/?randomnick=1&channels=youtube-dl)).
2122 Opening a bug report or suggestion
2124 Be sure to follow instructions provided below and in the issue tracker.
2125 Complete the appropriate issue template fully. Consider whether your
2126 problem is covered by an existing issue: if so, follow the discussion
2127 there. Avoid commenting on existing duplicate issues as such comments
2128 do not add to the discussion of the issue and are liable to be treated
2129 as spam.
2130 Please include the full output of youtube-dl when run with -v, i.e. add
2132 -v flag to your command line, copy the whole output and post it in the
2133 issue body wrapped in ``` for better formatting. It should look simi‐
2134 lar to this:
2135 $ youtube-dl -v <your command line>
2137 [debug] System config: []
2138 [debug] User config: []
2139 [debug] Command-line args: [u'-v', u'https://www.youtube.com/watch?v=BaW_jenozKcj']
2140 [debug] Encodings: locale cp1251, fs mbcs, out cp866, pref cp1251
2141 [debug] youtube-dl version 2015.12.06
2142 [debug] Git HEAD: 135392e
2143 [debug] Python version 2.6.6 - Windows-2003Server-5.2.3790-SP2
2144 [debug] exe versions: ffmpeg N-75573-g1d0487f, ffprobe N-75573-g1d0487f, rtmpdump 2.4
2145 [debug] Proxy map: {}
2146 ...
2147 Do not post screenshots of verbose logs; only plain text is acceptable.
2149 The output (including the first lines) contains important debugging in‐
2151 formation. Issues without the full output are often not reproducible
2152 and therefore do not get solved in short order, if ever.
2153 Finally please review your issue to avoid various common mistakes (you
2155 can and should use this as a checklist) listed below.
2156 Is the description of the issue itself sufficient?
2158 We often get issue reports that are hard to understand. To avoid sub‐
2159 sequent clarifications, and to assist participants who are not native
2160 English speakers, please elaborate on what feature you are requesting,
2161 or what bug you want to be fixed.
2162 Make sure that it’s obvious
2164 • What the problem is
2166 • How it could be fixed
2168 • How your proposed solution would look
2170 If your report is shorter than two lines, it is almost certainly miss‐
2172 ing some of these, which makes it hard for us to respond to it. We’re
2173 often too polite to close the issue outright, but the missing info
2174 makes misinterpretation likely. As a committer myself, I often get
2175 frustrated by these issues, since the only possible way for me to move
2176 forward on them is to ask for clarification over and over.
2177 For bug reports, this means that your report should contain the com‐
2179 plete output of youtube-dl when called with the -v flag. The error
2180 message you get for (most) bugs even says so, but you would not believe
2181 how many of our bug reports do not contain this information.
2182 If your server has multiple IPs or you suspect censorship, adding
2184 --call-home may be a good idea to get more diagnostics. If the error
2185 is ERROR: Unable to extract ... and you cannot reproduce it from multi‐
2186 ple countries, add --dump-pages (warning: this will yield a rather
2187 large output, redirect it to the file log.txt by adding >log.txt 2>&1
2188 to your command-line) or upload the .dump files you get when you add
2189 --write-pages somewhere (https://gist.github.com/).
2190 Site support requests must contain an example URL. An example URL is a
2192 URL you might want to download, like
2193 https://www.youtube.com/watch?v=BaW_jenozKc. There should be an obvi‐
2194 ous video present. Except under very special circumstances, the main
2195 page of a video service (e.g. https://www.youtube.com/) is not an exam‐
2196 ple URL.
2197 Is the issue already documented?
2199 Make sure that someone has not already opened the issue you’re trying
2200 to open. Search at the top of the window or browse the GitHub Issues
2201 (https://github.com/ytdl-org/youtube-dl/search?type=Issues) of this
2202 repository. Initially, at least, use the search term -label:duplicate
2203 to focus on active issues. If there is an issue, feel free to write
2204 something along the lines of “This affects me as well, with version
2205 2015.01.01. Here is some more information on the issue: ...”. While
2206 some issues may be old, a new post into them often spurs rapid activi‐
2207 ty.
2208 Are you using the latest version?
2210 Before reporting any issue, type youtube-dl -U. This should report
2211 that you’re up-to-date. About 20% of the reports we receive are al‐
2212 ready fixed, but people are using outdated versions. This goes for
2213 feature requests as well.
2214 Why are existing options not enough?
2216 Before requesting a new feature, please have a quick peek at the list
2217 of supported options (https://github.com/ytdl-org/youtube-dl/blob/mas‐
2218 ter/README.md#options). Many feature requests are for features that
2219 actually exist already! Please, absolutely do show off your work in
2220 the issue report and detail how the existing similar options do not
2221 solve your problem.
2222 Is there enough context in your bug report?
2224 People want to solve problems, and often think they do us a favor by
2225 breaking down their larger problems (e.g. wanting to skip already down‐
2226 loaded files) to a specific request (e.g. requesting us to look whether
2227 the file exists before downloading the info page). However, what often
2228 happens is that they break down the problem into two steps: One simple,
2229 and one impossible (or extremely complicated one).
2230 We are then presented with a very complicated request when the original
2232 problem could be solved far easier, e.g. by recording the downloaded
2233 video IDs in a separate file. To avoid this, you must include the
2234 greater context where it is non-obvious. In particular, every feature
2235 request that does not consist of adding support for a new site should
2236 contain a use case scenario that explains in what situation the missing
2237 feature would be useful.
2238 Does the issue involve one problem, and one problem only?
2240 Some of our users seem to think there is a limit of issues they can or
2241 should open. There is no limit of issues they can or should open.
2242 While it may seem appealing to be able to dump all your issues into one
2243 ticket, that means that someone who solves one of your issues cannot
2244 mark the issue as closed. Typically, reporting a bunch of issues leads
2245 to the ticket lingering since nobody wants to attack that behemoth, un‐
2246 til someone mercifully splits the issue into multiple ones.
2247 In particular, every site support request issue should only pertain to
2249 services at one site (generally under a common domain, but always using
2250 the same backend technology). Do not request support for vimeo user
2251 videos, White house podcasts, and Google Plus pages in the same issue.
2252 Also, make sure that you don’t post bug reports alongside feature re‐
2253 quests. As a rule of thumb, a feature request does not include outputs
2254 of youtube-dl that are not immediately related to the feature at hand.
2255 Do not post reports of a network error alongside the request for a new
2256 video service.
2257 Is anyone going to need the feature?
2259 Only post features that you (or an incapacitated friend you can person‐
2260 ally talk to) require. Do not post features because they seem like a
2261 good idea. If they are really useful, they will be requested by some‐
2262 one who requires them.
2263 Is your question about youtube-dl?
2265 It may sound strange, but some bug reports we receive are completely
2266 unrelated to youtube-dl and relate to a different, or even the re‐
2267 porter’s own, application. Please make sure that you are actually us‐
2268 ing youtube-dl. If you are using a UI for youtube-dl, report the bug
2269 to the maintainer of the actual application providing the UI. On the
2270 other hand, if your UI for youtube-dl fails in some way you believe is
2271 related to youtube-dl, by all means, go ahead and report the bug.
2272
2274 youtube-dl is released into the public domain by the copyright holders.
2275 This README file was originally written by Daniel Bolton
2277 (https://github.com/dbbolton) and is likewise released into the public
2278 domain.
2279 YOUTUBE-DL(1)