oneshotlm
Compare all models on this prompt →

Boids flocking · seed-2-1-turbo

Create a full-screen flocking simulation of boids using separation, alignment, and cohesion, drawn as oriented triangles that steer smoothly and flee from the mouse cursor.

⛨ sandboxed Open in new tab ↗
transcript · 4 messages
user

Create a full-screen flocking simulation of boids using separation, alignment, and cohesion, drawn as oriented triangles that steer smoothly and flee from the mouse cursor.

assistant
thinking
The user wants a full-screen boids flocking simulation with four specified behaviors.All content will be packed into one file, and I'll start building the simulation now.
write /tmp/ws/index.html (9256 bytes)
toolResult

Successfully wrote 9256 bytes to /tmp/ws/index.html

assistant

Created `/tmp/ws/index.html` — a full-screen boids flocking simulation with: - **Three core rules**: separation (avoid crowding), alignment (match heading), cohesion (steer toward group center) - **Oriented triangles** that point in the direction of movement, with glow effects and speed-based brightness - **Mouse flee behavior**: boids scatter from the cursor within a visible red radius, with falloff so closer boids flee harder - **Smooth steering**: velocity is clamped and forces are capped for fluid motion - **Screen wrapping**: boids wrap around edges seamlessly - **Trail effect**: semi-transparent overlay gives motion blur trails - **Touch support** and devicePixelRatio scaling for crisp rendering - Adaptive boid count based on screen size (~200 max) Open `index.html` in a browser to run it — no network needed.

TAKE IT TO CODE

A single self-contained HTML file — copy it and open in any browser, or reuse the prompt with your own pi-agent run.

RUN

modelseed-2-1-turbo
vendorByteDance · Seed
harnesspi-agent
generated2026-08-13

EVALUATION

4/5 working

Boids flock and flee smoothly from cursor with glowing trails, animating continuously and reacting to all inputs.

Vision grade · anthropic/claude-sonnet-5

MEASUREMENTS

Raw readings from the capture, not a verdict — the grade above is the only judgement. Percentages are how much of the screen changed; the measure under-reports sparse motion on a dark background.

idle motion19.7%
on click28.9%
on drag32.2%
on wheel28.0%
on enter + space24.6%
on arrow keys22.0%
on w/a/s/d21.0%
frame spread25.5 / 255
console errors0
js errors none

METRICS

runtime50.9s
tokens in6.9k
tokens out3.6k
cached2.7k
cost / run$0.0112