Banana Gun: trading tools rating breakdown
Messaging-app bot
One of the longest-running bots in this category, with a measured take rate of 0.0850% over thirty days — well below the terminals, and in the same range as what a decentralised exchange itself charges.
One of the longest-running bots in this category, with a measured take rate of 0.0850% over thirty days — well below the terminals, and in the same range as what a decentralised exchange itself charges.
How the score is built
Each criterion is scored 0–10 and weighted. The median column is the middle score across every entry in this ranking, so a row reads as a position rather than a number.
| Criterion | What we measured | Weight | Score | Category median | Rank | Weighted gap |
|---|---|---|---|---|---|---|
| The take rate, computed | 0.085 take rate · computed · 2026-08-27 · source | 35% | 8 | 4.5 | 3 of 8 | +1.23 |
| Who holds the key | — | 25% | 3 | 3 | 4 of 8 | 0.00 |
| Is the fee disclosed before the trade? | — | 20% | 6 | 6 | 2 of 8 | 0.00 |
| What the routing actually does | — | 12% | 6 | 7.5 | 6 of 8 | -0.18 |
| Chains and assets | — | 8% | 5 | 5 | 4 of 8 | 0.00 |
Measured 27 August 2026 · weights and method · decided by the take rate, computed, worth +1.23 points against the median
The take rate, computed: 8/10
Thirty days of fees collected divided by thirty days of volume routed, from public data. This is what the tool charges on top of whatever the underlying venue charges, and the spread is enormous: 0.0060% to 1.9287% across the tools measured here, against 0.0506% to 0.6639% at the decentralised exchanges they route into.
Scored 8 of 10 against a category median of 4.5, which places it 3rd of 8 among trading tools on this criterion. At a 35% weight that is 1.23 points above the median contribution of the weighted total. The best score in the category is 10, the worst 1.
Who holds the key: 3/10
Whether the tool generates and holds a private key on its own servers — the model almost every messaging-app bot uses — or signs from a wallet you control. A bot that holds the key is a custodian you cannot audit, with no attestation, no licence and no recovery process, and that fact belongs above execution speed in any honest ranking.
Scored 3 of 10 against a category median of 3, which places it 4th of 8 among trading tools on this criterion. At a 25% weight that is exactly level with the median of the weighted total. The best score in the category is 10, the worst 2.
Is the fee disclosed before the trade?: 6/10
Whether the tool states its fee plainly in its own documentation and shows it at the point of trade, or buries it in a percentage of an amount the user never sees itemised. We compare what is disclosed against the take rate we computed.
Scored 6 of 10 against a category median of 6, which places it 2nd of 8 among trading tools on this criterion. At a 20% weight that is exactly level with the median of the weighted total. The best score in the category is 9, the worst 4.
What the routing actually does: 6/10
Documented routing behaviour: which venues it can reach, whether it uses private transaction relays, and what protection it offers against sandwiching. Speed claims that cannot be verified from public data are not scored.
Scored 6 of 10 against a category median of 7.5, which places it 6th of 8 among trading tools on this criterion. At a 12% weight that is 0.18 points below the median contribution of the weighted total. The best score in the category is 9, the worst 6.
Chains and assets: 5/10
Chains supported with real routed volume, from public data at a stated timestamp.
Scored 5 of 10 against a category median of 5, which places it 4th of 8 among trading tools on this criterion. At a 8% weight that is exactly level with the median of the weighted total. The best score in the category is 9, the worst 3.
Its nearest neighbours in this ranking
Questions about this score
How does Banana Gun compare to other bots on price?
+
Our measurement puts it at 0.0850% of routed volume against Trojan's 1.1108% and GMGN's 0.9896% — an order of magnitude cheaper. The sample behind it is small, at $12.1m routed over thirty days, so treat it as an indication rather than a guarantee and check your own fills.