Switch branches/tags
Nothing to show
Find file
Fetching contributors…
Cannot retrieve contributors at this time
148 lines (100 sloc) 4.02 KB


A lightwight, dependency-free Python library for downloading YouTube Videos.

Downloading videos from YouTube shouldn't require some bloatware application, it's usually a niche condition you want to do so in the first place. So I Prsent to you, PyTube!


If you are on Mac OS X or Linux, chances are that one of the following two commands will work for you:

$ easy_install pytube

or even better:

$ pip install pytube

or you can get the source code from github.


The only features I see implementing in the near future are:

  • Allow it to run as a command-line utility.
  • Making it compatible with Python 3.

Usage Example

from pytube import YouTube

# not necessary, just for demo purposes
from pprint import pprint

yt = YouTube()

# Set the video URL.
yt.url = ""

# Once set, you can see all the codec and quality options YouTube has made
# available for the perticular video by printing videos.


#[<Video: MPEG-4 Visual (.3gp) - 144p>,
# <Video: MPEG-4 Visual (.3gp) - 240p>,
# <Video: Sorenson H.263 (.flv) - 240p>,
# <Video: H.264 (.flv) - 360p>,
# <Video: H.264 (.flv) - 480p>,
# <Video: H.264 (.mp4) - 360p>,
# <Video: H.264 (.mp4) - 720p>,
# <Video: VP8 (.webm) - 360p>,
# <Video: VP8 (.webm) - 480p>]

# The filename is automatically generated based on the video title.
# You can override this by manually setting the filename.

# view the auto generated filename:
print yt.filename

#Pulp Fiction - Dancing Scene [HD]

# set the filename:
yt.filename = 'Dancing Scene from Pulp Fiction'

# You can also filter the criteria by filetype.


#[<Video: Sorenson H.263 (.flv) - 240p>,
# <Video: H.264 (.flv) - 360p>,
# <Video: H.264 (.flv) - 480p>]

# and by resolution

#[<Video: H.264 (.flv) - 480p>, 
#<Video: VP8 (.webm) - 480p>]

# to select a video by a specific resolution and filetype you can use the get
# method.

video = yt.get('mp4', '720p')

# NOTE: get() can only be used if and only if one object matches your criteria.
# for example:


#[<Video: MPEG-4 Visual (.3gp) - 144p>,
# <Video: MPEG-4 Visual (.3gp) - 240p>,
# <Video: Sorenson H.263 (.flv) - 240p>,
# <Video: H.264 (.flv) - 360p>,
# <Video: H.264 (.flv) - 480p>,
# <Video: H.264 (.mp4) - 360p>,
# <Video: H.264 (.mp4) - 720p>,
# <Video: VP8 (.webm) - 360p>,
# <Video: VP8 (.webm) - 480p>]

# Notice we have two H.264 (.mp4) available to us.. now if we try to call get()
# on mp4..

video = yt.get('mp4')
# MultipleObjectsReturned: get() returned more than one object -- it returned 2!

# In this case, we'll need to specify both the codec (mp4) and resolution
# (either 360p or 720p).

# We can also get the highest resolution available using the 
# get_highest_quality() method. An optional preferred filetype extension can be 
# specified as a string, otherwise we default to 'mp4', if available. The function
# will favour the highest quality over the preferred filetype.
video = yt.get_highest_res(preferred='mp4')

# Okay, let's download it!

# Downloading: Pulp Fiction - Dancing Scene.mp4 Bytes: 37561829
# 37561829  [100.00%]

# Note: If you wanted to choose the output directory, simply pass it as an 
# argument to the download method.'/tmp/')


After missing the deadline to register for PyCon 2012, I decided to write what became PyTube and crawler to collect all the YouTube links for the talks on

To avoid having to encode them to mp4 (so I could watch them on my iPhone) I wrote it so you could specify an encoding format.

In recently weeks interest has picked up in the project, so I decided to dedicate more time to further its development and actively maintain it.


My only real goal for this is to never require any third party dependancies, to keep it simple and make it reliable.