They were pretty good by the 1840s for the Sikh Wars!
If they are native levies led as evilgong notes by political figures, and armed with blades or – at best – obsolete firearms and not very many of those, they're not going to do well against troops led and armed European-style. The raw material of native troops and sepoys was presumably much the same, with leadership, equipment and organisation being the differentiators.
In modern management-speak, it's all about People, Process and Technology. At Rorke's Drift 100 guys with rifles held 4,000 Zulus with (mostly) spears to a draw. If you look at what either side needed to turn the draw into a victory, the defenders needed a piece of Technology, eg machine-guns or something, to get beyond standing the Zulus off and move up to scything them down at a rate that would rout them.
The Zulus for their part needed a Process to break into a fortified position defended by rifles; once inside the perimeter, both sides are armed with spears, in effect. Nobody really had such a process until about 1918.
Giving either side another 100 guys armed with spears wouldn't have made any difference. Only having spears wasn't the big Zulu problem. European cavalry mostly used spears too and tactically must have manoeuvred about as fast as Zulu light infantry did.
Indian wars, same thing. Mahratta armies – lots of People but a lack of Process and Technology; European armies – not many people, compensated for by more Process and more Technology.
Of those, probably the Process is most significant. Mobs of disordered troops tend to lose to formed bodies of well-ordered troops even when the latter have the same technology and numbers.