Honest comparison across the categories people mistake us for: AI trading bots, copy-trading platforms, DeFi yield products. Where we win, where we don't, where we're just differently shaped.
AIQuant (AI Quant Labs, Atlanta) went GA Sep 2025 with multi-agent across Solana + BNB + Base. 103,500+ trades, ~$25M volume pre-GA, $AIQ token listed. Ships the closest shape to Kabal. Their gaps are our wedges.
| AIQuant.fun | Kabal | |
|---|---|---|
| Architecture | Customizable multi-agent (generic) | 8 named agents · typed event bus |
| Rug filter | Not documented | Ghost v1 — 5 pre-trade checks |
| On-chain memo proof | No | Yes — SPL Memo on every trade |
| Reasoning transparency | Not documented | Agent role cards · knowledge base · Council |
| Per-agent paper/live A/B | No | Yes — unique to Kabal |
| Risk layers | 1 (strategy-level) | 3 (per-trader · Sentinel · Aegis) |
| Chain focus | Multi-chain generic (Solana + BNB + Base) | Solana-memecoin specialist |
| Monetization | Pay-per-agent hatching fee ($AIQ token) | Session-based SOL pricing + perf-linked pause |
| Learning loop | Not disclosed | Librarian + 42-wallet labeled losing cohort |
| Kill switch | Not documented | /halt stops all positions in <1s |
TrenchFeed has the closest product philosophy (non-custodial, filter- first, 3 modes Sniper/Default/Careful, 0.95 SOL/wk Premium Plus). Strong filter shop, weak executor. We don't combat — we co-position.
| TrenchFeed | Kabal | |
|---|---|---|
| Shape | 3 scanner modes + bolt-on exec | 8-agent squad + ghost + memo + librarian |
| Rug filter | Ghost scoring | Ghost v1 · 5 filters, provenance documented |
| Agents | None (3 modes only) | 8 named agents with role cards + knowledge |
| Per-agent A/B | No | Yes — paper + live same process |
| On-chain memo | No | SPL Memo per live trade |
| Pricing (weekly) | 0.95 SOL/wk (Premium Plus) | 2 SOL/wk (justified by squad depth) |
| Learning loop | Static heuristics | Librarian + 42-wallet labeled cohort |
| Telegram bots | Kabal | |
|---|---|---|
| Role | Fastest execution terminal | Agent brain above the terminal |
| Latency | Sub-3s (Axiom owns 60-72% share) | Not competing on latency — routing through KKM |
| Reach | ~5M users combined | Kabal plugs into that via KKM copytrading (free tier) |
| Discipline layer | Pay-per-trade, no caps, no ghost, no library | Ghost v1 · 3-tier risk · kill-switch <1s · memo proof |
| Squad | One bot, one behavior | 8 agents with distinct roles and strategies |
| Lending / LP | Kabal Vault | |
|---|---|---|
| Yield source | Stablecoin interest / IL | Active trading PnL |
| Typical APR | 4-12% | 12-200% (tier-dependent, variable) |
| Drawdown exposure | Low (or depeg) | Medium-high (capped per tier) |
| IL exposure | Yes (LP) / No (lend) | No — single-asset exposure |
| Transparency | On-chain state only | On-chain state + memo trail of decisions |
| Strategy change | Governance vote or protocol upgrade | Trader's discretion + Sentinel, memo'd |
| Best for | Ultra-conservative passive carry | Active exposure with discipline rails |
Four categories where the competition has us for now, and what we're doing about it:
1. Distribution reach — Trojan 2M users, Axiom ~60-72% Solana terminal share, BONKbot 526K users. Our answer: route through KKM (free copy-trading tier) rather than fight for Telegram distribution head-on.
2. Token / multi-chain story — AIQuant has $AIQ listed + 3 chains deployed. We don't have a token yet (by design) and we're Solana-only through 2027. Our answer: the moat is the discipline layer, not the reach.
3. Brand recognition — We have 294 paper trades in one baseline. Time + the V.1 live cutover fix this, not marketing.
4. External audit — Not yet audited. V.0 is paper. Cold-start caps (0.02 SOL) structurally limit live exposure during burn-in. External audit planned Q3 2026.
This page is deliberately specific, not diplomatic. If you think we got a comparison wrong, tell us — press@kabal.trade. We'd rather fix an inaccuracy than defend a bad claim.