Word embedding created with gensim based on Swedish tweets from 2016. Script used for producing word embeddings is on github.
The model is trained on tweets in Swedish, English, Norwegian and Finnish mainly.
To load the model you need to run Python 3.4 (tested with 3.4.6) due to a quirk in how Gensim works.
model = gensim.models.Word2Vec.load("big_model_150_mincount_30_no_stops")