Aryabhatta is usually credited to have invented the number zero, at the least in India. One of the arguments for it is that he knew the place value system of the 10s, 100s, 1000s, etc. But this argument takes the use of zero back to 1500-1200 BCE. In the white Yajurveda (17.2), this very system is mentioned upto 12 zeros. It is Brahmagupta’s book on math, Brahmasphutasiddhanta (kuttukadhyaya, 30), where we see an obvious mention of zero (shunya) being used as a number to describe the difference between, for example, +2 & -2. This is back in 7th century CE.
I wanted to find a similar text where zero or shunya had been used as a specific number or digit. I then found one book, dated around 200-100 BCE, called the Chhanda-Sutra or Chhanda-Shastra, written by Pingala. It is a book on sanskrit prosody. It describes various precise calculations regarding precise poetic meters. In its sutra numbers 28 & 29 of chapter 8, Pingala mentions a certain calculation for which he uses two numbers as place markers. The numbers he uses are two (dvi) & zero (shunya). His formula is based on halving a number. For example, if you have to halve number 6, you should write ‘2’ next to it signifying that it has been divided into ‘two’ parts. Then minus 3 from 6 (its half), and you get 3. As 3 is an odd number, and as the author suggests to subtract 1 from it to make it even, it is not halved yet. Hence ‘o’ is written next to it signifying that number 3 has been divided into zero parts (no parts). The next calculation is not related to this discussion.
I first had a doubt about the usage of the word ‘shunya’ in the sutras. But the single most point that it has been used along with the number ‘two’, signifying into how many parts is a number divided (2 for two parts, 0 for no parts), I am pretty sure that zero has been regarded as a separate number by Pingala back in the 2nd or 1st century BCE. It might not be necessarily true that Pingala invented the number ‘shunya’. But it is possible that he knew it as a separate number.