1: \begin{abstract}

2: This paper is devoted to extend some  Hu-Keel results on Mori dream spaces (MDS) beyond the projective setup. Namely, $\Q$-factorial algebraic varieties with finitely generated class group and Cox ring, here called \emph{weak} Mori dream spaces (wMDS), are considered.  Conditions guaranteeing the existence of a neat embedding of a (completion of a) wMDS into a  complete toric variety are studied, showing that, on the one hand, those which are complete and admitting low Picard number are always projective, hence Mori dream spaces in the sense of Hu-Keel. On the other hand, an example of a wMDS that does not admit any neat embedded \emph{sharp} completion (i.e. Picard number preserving) into a  complete toric variety is given, on the contrary of what Hu and Keel exhibited for a MDS.

3: 

4: Moreover, termination of the Mori minimal model program (MMP) for every divisor and a classification of rational contractions for a complete wMDS are studied, obtaining analogous conclusions as for a MDS.

5: 

6: Finally, we give a characterization of wMDS arising from a small $\Q$-factorial modification of a projective weak $\Q$-Fano variety.

7: 

8: \keywords{Mori dream space \and  Cox ring \and  Class group \and  toric varieties \and  Gale duality \and  the secondary fan \and  GKZ decomposition \and  good and geometric quotient \and  fan matrix \and  weight matrix \and  nef cone \and  moving cone \and  pseudo-effective cone \and  Picard number \and  bunch of cones \and  irrelevant ideal and locus \and  completion \and  completion of fans \and  minimal model program \and  small modification \and  rational contraction}

9: % \PACS{PACS code1 \and PACS code2 \and more}

10: \subclass{14E25 \and 14M25 \and 14C20 \and 14E30 \and 14L24 \and 14L30}

11: \end{abstract}

12: