Get highest unique number or NULL if none exists

I want biggest possible single number from the table. There are duplicates in the table so I want the final result to reflect null because there is no single biggest number.

I tried every possible modification to my query to have my code return null values but it doesn't. This is leetcode problem that I was trying to solve and trying to build a much deeper understanding of the problem in my SQL query.

Just needed to know how can I introduce NULL values manually. I know the actual coding answer to the question which is why I need some good explanation of what I am doing wrong along with modification to this code.

Thank you for all the help you provide.

Input:

Output:

Expected:

Here is the code:

/* Write your T-SQL query statement below */
WITH CTE AS
(
    SELECT *,
    ROW_NUMBER() OVER (PARTITION BY num ORDER BY num) AS RN
    FROM MyNumbers
),
CTE2 AS
(
    SELECT TOP 1 num 
    FROM CTE
    WHERE num NOT IN (SELECT num FROM CTE WHERE RN >= 2)
    ORDER BY num DESC
)
SELECT 
    (CASE
         WHEN num IS NULL OR num = '' 
             THEN NULL
             ELSE num
     END) AS num FROM CTE2;

With your approach, using row_number()

WITH CTE AS
(
    SELECT *,
    ROW_NUMBER() OVER (PARTITION BY num ORDER BY num) AS RN
    FROM MyNumbers
),
CTE2 AS
(
    SELECT TOP 1 num 
    FROM CTE
    WHERE num NOT IN (SELECT num FROM CTE WHERE RN >= 2)
    ORDER BY num DESC
)
SELECT 
    max(CASE
         WHEN num IS NULL OR num = '' 
             THEN NULL
             ELSE num
     END) AS num FROM CTE2;

If CTE2 not returns rows, max(...) returns null.
See example with empty table

create table test3(num int);
select max(num) maxNum from test3;

fiddle

If any case, CTE2 returns rows or not, you should calculate max(num).

Next example:
You can use count(*)over(...) and filter rows with count=1

WITH CTE AS
(
    SELECT *,
      count(*) OVER (PARTITION BY num ORDER BY num) AS cnt
    FROM test MyNumbers
),
CTE2 AS
(
    SELECT num 
    FROM CTE
    WHERE cnt=1
)
select max(num) maxNum from cte2

It is possible to simplify, moving filter (WHERE) to main query

WITH CTE AS
(
    SELECT *,
    ROW_NUMBER() OVER (PARTITION BY num ORDER BY num) AS RN
    FROM test MyNumbers
)
SELECT max(num) AS num 
FROM CTE
WHERE num NOT IN (SELECT num FROM CTE WHERE RN >= 2);

Or

WITH CTE AS
(
    SELECT *,
      count(*) OVER (PARTITION BY num ORDER BY num) AS cnt
    FROM test MyNumbers
)
select max(num) maxNum 
from cte
WHERE cnt=1

Latest:
directly count and filter rows by HAVING clause

select max(numg) num
from( select num numg from test group by num having count(*)=1 )a

Fiddle