Jupiter volume bot: the convenience that costs the most
Jupiter is not a venue. It is a router that finds a path across the venues, and for a person making one swap that is straightforwardly useful. For a fleet making thousands of swaps it is the most expensive routing decision available, and our block measurements show why: aggregated routes produce larger transactions touching more accounts, and larger transactions lose the contention race more often. Every one of those losses is charged in full.
What an aggregator actually does to your transaction
The routing itself is genuinely good. If a token has liquidity scattered across several pools, an aggregator will usually find better execution than any single pool provides, and for large trades that price improvement is real money. The question for a campaign is whether the improvement covers the additional failure risk, and at the trade sizes a volume campaign uses, it usually does not.
A volume campaign makes many small trades. Small trades barely benefit from route optimisation because price impact is negligible either way. What they are sensitive to is the per-transaction cost of landing at all, and that is exactly what aggregation makes worse.
Measured: the highest failure rate we sample
| Program | Sampled tx | Failure rate | Median fee | Cost per landed swap |
|---|---|---|---|---|
| Jupiter aggregator v6 | 286 | 55.9% | 9,097 | 20,628 |
| Pump.fun AMM (post-graduation) | 2,341 | 23.3% | 8,243 | 10,747 |
| Meteora DLMM | 863 | 38.9% | 6,642 | 10,871 |
| Orca Whirlpools low sample | 35 | 31.4% | 5,432 | 7,918 |
| Raydium CLMM | 54 | 48.1% | 5,081 | 9,790 |
This is one of the findings we did not expect when the sampler was first written, and it is the reason the routing control in the console defaults to direct rather than aggregated. The fee per attempt on aggregator routing is not dramatically higher. The share of attempts that produce nothing is.
Two mechanisms explain most of it. First, a transaction referencing more writable accounts collides with more other transactions competing for the same accounts, and the loser is dropped. Second, aggregated routes carry tighter execution constraints, so a route computed a moment ago can fail its own slippage check by the time it lands. Both effects scale with how busy the venue is, which is why the gap widens during launch windows.
Method, sample size and slot range are published on the measurement page. Every figure here is read from finalized blocks rather than estimated.
When routing through an aggregator is still right
| Situation | Better path | Why |
|---|---|---|
| One deep pool holds the liquidity | Direct | Nothing to aggregate; you pay the failure premium for no benefit |
| Liquidity split across several pools | Aggregated | Route improvement can exceed the failure premium |
| Many small trades | Direct | Price impact is negligible at small size, so there is nothing to improve |
| Few large trades | Aggregated | Slippage saved on size outweighs the extra failures |
| Contested launch window | Direct | Contention is exactly when big transactions lose most often |
Volion resolves the deepest routing target from the mint and routes directly to it by default, with aggregation available as an explicit choice rather than a silent one. The console shows which program it resolved to and prices the campaign against that program measured failure rate, so switching routing changes the estimate you see before funding.
Does routing through Jupiter change what aggregators show
This matters because it removes the only argument for paying the failure premium on a volume campaign. If aggregated routing produced better indexing, the extra cost would buy something. It does not. The pool records a trade, the indexer picks it up, and the transaction path that produced it is not part of what gets counted.
What does affect how activity is read is the shape of it: how many distinct wallets traded, how spread out the trades were, and whether the sizes look chosen by people or generated by a loop. What aggregators index and how fleets get identified both go into that in detail.
For the venue-by-venue picture, the Solana volume bot overview links every routing target we measure, and the pricing page shows what each one costs at current rates.