qwen3.5:9b vs qwen3:14b

Both models answered the same one-shot tasks under identical rules — no retries, answers published unedited. Two separate axes, never blended: does the code actually run, and how do blind-scored text answers hold up against a written gold answer.

Model	Runnable sims	Blind text avg	Failure modes
qwen3.5:9b	36% (31/85)	9.21 (38 scored)	ERR 26 · STAT 11 · BLANK 16 · HANG 1
qwen3:14b	28% (24/85)	8.71 (38 scored)	ERR 39 · STAT 10 · BLANK 12

Biggest contrasts

Rounds where the two models diverged most — open them and read both answers side by side.

Round	qwen3.5:9b	qwen3:14b
R140 · Water sloshing in a tilting container	OK	ERR
R126 · Bubble sort sweeping bars into order	OK	ERR
R124 · Dijkstra's shortest path over weighted terrain	STAT	OK
R120 · Scanline fill of a self-intersecting polygon	ERR	OK
R119 · Floyd-Steinberg dithering a gradient	OK	ERR

→ Interactive round-by-round comparison · → Full leaderboard

Generated 2026-07-31 from the published rounds. Harness failures are excluded from every base — a broken cluster node is our fault, not the model's.