One-shot output, unedited

The community's one-shot vibe checks, generated by a model running on one desktop GPU. Every artifact below is exactly what came back from a single request — nothing was fixed by hand, and a file that hit the token limit is marked failed rather than quietly shown.

10 runs · 9 usable · configurations: f16 KV (control), q4_0 KV (daily profile)

Flappy Bird, one shot

runsf16 KV (control) · seed 42 · 7380 tok · 53.6s · 137.6 tok/s
The classic zero-shot game test: game loop, collision, scoring and restart, with no follow-up turn to fix anything.

Flappy Bird, one shot

runsq4_0 KV (daily profile) · seed 42 · 7772 tok · 59.0s · 131.8 tok/s
The classic zero-shot game test: game loop, collision, scoring and restart, with no follow-up turn to fix anything.
major
Difficulty tuning makes the first gap effectively unpassable: PIPE_SPACING is 220px on a canvas capped at 440px, so two pipes are always on screen with 154px of clear air between them.observed + code
latent
Frame-rate independence is plumbed but dead. loop() computes dt = (t-last)/16.667, clamps it, and passes it to update(dt) - which never multiplies by it. Physics runs per-frame, not per-second.Harmless on the 60Hz display it was tested on; a 144Hz monitor runs the game 2.4x too fast.code: p.x -= PIPE_SPEED; bird.vy += GRAVITY; bird.y += bird.vy;

Balls in a spinning heptagon

runsf16 KV (control) · seed 42 · 3775 tok · 27.5s · 137.3 tok/s
The viral physics one-shot. Needs collision detection against rotating walls, gravity, damping and a render loop - all in one pass with no chance to debug.

Balls in a spinning heptagon

failedq4_0 KV (daily profile) · seed 42 · 14000 tok · 99.7s · 140.4 tok/s
The viral physics one-shot. Needs collision detection against rotating walls, gravity, damping and a render loop - all in one pass with no chance to debug.
fatal
Never completed - hit the 14,000 token ceiling mid-function.finish_reason: length
truncated: hit max_tokens

Zoomable Mandelbrot

runsf16 KV (control) · seed 42 · 7642 tok · 56.5s · 135.2 tok/s
Pure numerical code plus an interaction model. Wrong escape-time maths produces something that still renders but is visibly not a Mandelbrot set, so it fails loudly rather than silently.

Zoomable Mandelbrot

runsq4_0 KV (daily profile) · seed 42 · 6196 tok · 47.1s · 131.4 tok/s
Pure numerical code plus an interaction model. Wrong escape-time maths produces something that still renders but is visibly not a Mandelbrot set, so it fails loudly rather than silently.
fatal
Renders a black screen.observed
fatal
HiDPI mismatch: createImageData(W,H) allocates in CSS pixels while the canvas backing store is W*DPR x H*DPR. putImageData ignores the current transform and writes in device pixels, so on a DPR=2 display the image lands in the top-left quarter and the rest stays blank.code: canvas.width = W*DPR (line 265) vs img = ctx.createImageData(W,H) (line 270)

Pelican on a bicycle

runsf16 KV (control) · seed 42 · 2235 tok · 16.1s · 139.1 tok/s
Simon Willison's benchmark. A scene that cannot exist in the training data, so the model has to compose it: valid SVG syntax, bicycle structure, pelican anatomy, and the spatial relation between them.

Pelican on a bicycle

runsq4_0 KV (daily profile) · seed 42 · 2026 tok · 14.6s · 139.1 tok/s
Simon Willison's benchmark. A scene that cannot exist in the training data, so the model has to compose it: valid SVG syntax, bicycle structure, pelican anatomy, and the spatial relation between them.
major
Pelican's legs reach the pedals but then extend upward past the top of the viewBox; the figure never terminates.observed
note
Static by design - the benchmark asks for a still SVG, not an animation.by design

Solar system

runsf16 KV (control) · seed 42 · 7722 tok · 57.6s · 134.1 tok/s
Tests whether the model holds a set of real facts (relative orbital periods and sizes) while writing an animation, rather than picking one or the other.

Solar system

runsq4_0 KV (daily profile) · seed 42 · 7656 tok · 58.6s · 130.6 tok/s
Tests whether the model holds a set of real facts (relative orbital periods and sizes) while writing an animation, rather than picking one or the other.
none
Orbital periods verified against real values: all eight planets match to three decimal places, 0% deviation relative to Earth.checked programmatically