It is clear when doing

 data.groupby(['A','B']).mean()

We get something multiindex by level 'A' and 'B' and one column with the mean of each group

how could I have the count(), std() simultaneously ?

so result looks like in a dataframe

A   B    mean   count   std

The following should work:

data.groupby(['A','B']).agg([pd.Series.mean, pd.Series.std, pd.Series.count])

basically call agg and passing a list of functions will generate multiple columns with those functions applied.

Example:

In [12]:df = pd.DataFrame({'a':np.random.randn(5), 'b':[0,0,1,1,2]})
df.groupby(['b']).agg([pd.Series.mean, pd.Series.std, pd.Series.count])
Out[12]:a                mean       std count
b                          
0 -0.769198  0.158049     2
1  0.247708  0.743606     2
2 -0.312705       NaN     1

You can also pass the string of the method names, the common ones work, some of the more obscure ones don't I can't remember which but in this case they work fine, thanks to @ajcr for the suggestion:

In [16]:
df = pd.DataFrame({'a':np.random.randn(5), 'b':[0,0,1,1,2]})
df.groupby(['b']).agg(['mean', 'std', 'count'])Out[16]:a                mean       std count
b                          
0 -1.037301  0.790498     2
1 -0.495549  0.748858     2
2 -0.644818       NaN     1