One year from dorm-room drones to a jet UAV prototype
The speed ladder: slow drones are easy kills
Shahed-class loitering munitions cruise at 150–200 km/h — slow enough for modern air defenses to intercept at rates the video puts above 75%. Apollyon's bet is that at 700 km/h, nothing short of a multi-million-dollar surface-to-air missile can touch you.
View as table
| Platform | Speed (km/h) | Noted in video as |
|---|---|---|
| Shahed-class loitering munition | 150–200 | “barely 150 to 200 kmph” |
| Indian Army 10-inch FPV drones | ≈150–200 | “a little slower or on the same lines” |
| 5-inch FPV kamikaze (demoed) | ≈200 | high thrust-to-weight config |
| Apollyon kamikaze drone (original) | 300 | what the company started with |
| Apollyon interceptor (new target) | 350–400 | up from ~300 on earlier ADI design |
| Nightshade ADX-1 jet UAV | 700 | “cruise-missile-class tech for the UAV segment” |
Speed, payload, and reach — the jet UAV vs. what came before
Traditional FPV strike drones hit targets 10–30 km away with 1–1.5 kg of explosive. The Nightshade is pitched as a different class: 5 kg of payload delivered 100+ km deep, at 3–4× the speed.
View as table
| Measure | Traditional strike drones | Nightshade ADX-1 |
|---|---|---|
| Speed (km/h) | 150–300 | 700 |
| Payload (kg) | 1–1.5 | 5 |
| Strike range (km) | 10–30 | 100+ |
“Even if you intercept every Shahed, you still lose”
The core problem Apollyon says it's solving: cost asymmetry. Shooting down a ~$50K drone with a $1–1.5M missile means the defender bleeds money on every successful intercept. An aggressor can keep sending cheap drones until the defender goes broke.
View as table
| Asset | Approx. cost |
|---|---|
| Shahed-class attack drone | $50,000 |
| Surface-to-air missile to stop it | $1,000,000–1,500,000 |
The catch-22 the Nightshade creates
Flip the same logic around: when a fast, deep-striking jet UAV is inbound, the defender has only two moves — and both cost them.
Over 75% of traditional loitering munitions get intercepted
In Ukraine, combined soft-kill (jamming) and hard-kill (physical interception) rates against conventional loitering munitions exceed 75% in some cases. Each square below is one drone in a wave of 100.
Soft kill vs. hard kill
“I thought a kill is a kill.” Not quite — the two families of counter-drone defense fail in very different ways.
Soft kill — electronic warfare
Jammers and spoofers attack the drone's radio link or GPS instead of the drone itself. Cheap per engagement, but not reliable: against a drone with EW hardening (like frequency hopping), jamming “will do nothing.”
Hard kill — physical destruction
An interceptor rams the target (kinetic kill) or detonates a small warhead next to it (proximity fuse). More decisive: “when you're blowing something up, it can't keep coming at you.”
Radar guides the flight; a seeker finishes the job
The production interceptor engages in phases: a ground radar steers it for the initial and mid-course legs, then an onboard seeker takes over for the last ~500 m — fully automated, no link required.
Kinetic-kill variant
Rams the target to break its fuselage — used against smaller loitering munitions. The airframe itself is the weapon.
Explosive variant
Carries a small warhead with a proximity fuse — detonates near the target rather than needing a direct hit. “That is a 100% kill rate, basically.”
Frequency hopping: why their drones survived army jammers
A jammer floods one slice of spectrum with noise. FHSS (frequency-hopping spread spectrum) moves the control link across frequencies ~400 times per second — the jammer only ever blankets a sliver of the transmission, so the link holds.
A 20:1 thrust-to-weight ratio
The racing drone flown in the demo weighs about 500 g but its motors produce 10 kg of thrust — a thrust-to-weight ratio of 20:1. (A modern fighter jet manages roughly 1:1.) That's what buys ~200 km/h speed and tornado-tight maneuvers — at the price of a ~2-minute battery.
One airframe, two mission families
ADX-1 stands for “Apollyon Dynamics Experiment 1” — the first prototype, running German and Taiwanese jet engines, about a year from field deployment. The production system is designed around two roles.
Deep strike
SEAD — suppression of enemy air defenses. Home in on radars and SAM batteries and destroy them: “a blind enemy is an enemy that cannot strike.” Without working air defenses, the enemy can't contest the sky without massive fighter losses.
Runway denial (offensive counter-air). Crater enemy runways preemptively at the start of a war — “the fanciest fifth-generation fighter jets are just fancy toys on the ground” if they can't take off.
Target / decoy
Training target. Fitted with radar cross-section enhancers, the small UAV appears on an air-defense radar as a full-sized manned aircraft — a high-speed expendable aerial target for SAM crews to practice on.
Decoy. The same aircraft sent over enemy territory baits air defenses into wasting $1M+ surface-to-air missiles on an inexpensive UAV — draining the defender's magazine and wallet.