Qwen 3.6 27B NVFP4 Instruct | Benchmark #197

Described in:Asteroids-like game built by Qwen 3.6 and Qwen 3.8 as agents

Benchmark summary

Asteroids-like game built in SC# as an agent

ModelQwen 3.6 27B QuantizationNVFP4 ModeInstruct

The model builds an Asteroids-like space rocks shooter with power-ups in SC# as an agent: it writes the app, compiles it, runs its own tests and fixes the code. Hidden tests check the rules from the task; each hidden test is its own result.

Score 15/15 100.0% passed
15 passed15 total 2 AI queries 25 tool calls TOKENS IN1 142 158 TOKENS OUT17 162

Results by difficulty

1 group
Level Unknown / other
15 / 15 100.0%
passed
2 AI queries 25 tool calls IN1 142 158 OUT17 162

Results by category

1 group
Category Other
100.0% 15/15 passed
2 AI queries 25 tool calls TOKENS IN1 142 158 TOKENS OUT17 162

Tests with multiple category tags are counted in each matching category, so category totals may exceed the report total.

Benchmark items
Script / test Status Tokens
Building Games in SC#: Asteroids-like space rocks shooter 15 passed
100.0%
IN1 142 158 OUT17 162
✓
AsteroidsThe model builds an Asteroids game with power-ups from the task as an agent; this item passes when the app compiles and can be exported, and each hidden test is its own item.2 AI queries 25 tool calls
Passed IN1 142 158 OUT17 162
✓
Hidden_FiringSpace fires a bullet from the ship's nose; it flies at 6 px per frame, lives 40 frames, and the next shot comes 8 frames later; at most 4 bullets exist.0 AI queries 0 tool callsHiddenTest
Passed IN0 OUT0
✓
Hidden_HitDistanceA bullet hits a rock only when its center is closer than the rock's radius.0 AI queries 0 tool callsHiddenTest
Passed IN0 OUT0
✓
Hidden_InitialStateAfter AppInit the ship, the counters and the rocks of wave 1 have their starting values.0 AI queries 0 tool callsHiddenTest
Passed IN0 OUT0
✓
Hidden_NextWaveWhen no rock is left, the next wave starts with one large rock more, at least 60 px from the ship.0 AI queries 0 tool callsHiddenTest
Passed IN0 OUT0
✓
Hidden_PowerUpsAppearAt least one power-up appears for every 10 destroyed rocks, at the position of a destroyed rock.0 AI queries 0 tool callsHiddenTest
Passed IN0 OUT0
✓
Hidden_PowerUpsExpirePower-ups that are not picked up disappear after 300 frames.0 AI queries 0 tool callsHiddenTest
Passed IN0 OUT0
✓
Hidden_RapidFireAndShieldRapid fire shortens the cooldown to 3 frames, and the shield protects the ship from rocks.0 AI queries 0 tool callsHiddenTest
Passed IN0 OUT0
✓
Hidden_RocksSplitA bullet splits a large rock into two medium rocks at its position, a medium rock into two small ones; small rocks vanish; points are 20, 50 and 100.0 AI queries 0 tool callsHiddenTest
Passed IN0 OUT0
✓
Hidden_ScreenShowsTheFieldThe screen shows the rocks, a power-up, a bullet, the ship and the texts.0 AI queries 0 tool callsHiddenTest
Passed IN0 OUT0
✓
Hidden_ShipHitA rock that touches the ship costs a life; the ship goes back to the center and is safe for 90 frames; without lives the game is over.0 AI queries 0 tool callsHiddenTest
Passed IN0 OUT0
✓
Hidden_ThrustAndDriftUp pushes the ship by 0.15 per frame in its heading up to a speed of 4; without Up the velocity shrinks by 1 % per frame.0 AI queries 0 tool callsHiddenTest
Passed IN0 OUT0
✓
Hidden_TripleShotTriple shot fires three bullets 15 degrees apart for 300 frames.0 AI queries 0 tool callsHiddenTest
Passed IN0 OUT0
✓
Hidden_TurningLeft and Right turn the ship by 6 degrees per frame and keep the angle between 0 and 360.0 AI queries 0 tool callsHiddenTest
Passed IN0 OUT0
✓
Hidden_WrappingThe ship wraps around the edges of the screen.0 AI queries 0 tool callsHiddenTest
Passed IN0 OUT0