Skip to main content

← Dracon Bench leaderboard

Head-to-head

mistral-large-4-1t-a52b vs nemotron-3-nano-omni-30b-a3b-reasoning

The same frozen prompts at temperature 0, scored the same way. Every figure carries the items it was taken over, and the delta is always mistral-large-4-1t-a52b minus nemotron-3-nano-omni-30b-a3b-reasoning.

Scores

mistral-large-4-1t-a52b against nemotron-3-nano-omni-30b-a3b-reasoning.
Suitemistral-large-4-1t-a52bnemotron-3-nano-omni-30b-a3b-reasoningDelta
Overall46.0% 5 of 6· #2163.8% 6 of 6· #12-17.8 pts
IFEval-50 prompt80.0% n=10 49.0%–94.3%90.0% n=10 59.6%–98.2%-10.0 pts
IFBench-50 prompt0.0% n=10 0.0%–27.8%60.0% n=10 31.3%–83.2%-60.0 pts
MMLU-Pro-5010.0% n=10 1.8%–40.4%20.0% n=10 5.7%–51.0%-10.0 pts
JSON briefs80.0% n=5 37.6%–96.4%100.0% n=5 56.6%–100.0%-20.0 pts
GSM8K-5060.0% n=10 31.3%–83.2%80.0% n=10 49.0%–94.3%-20.0 pts
SVG totaln/a66/200 blind rubric

Deltas compare scores, not denominators: a 10-point gap means less when one side answered far fewer items. The n= lines keep that honest.

beach

Sun low over water with a reflection path; distant objects smaller and hazier than near ones.

mistral-large-4-1t-a52b

unjudged

no svg

nemotron-3-nano-omni-30b-a3b-reasoning

18/50

clock

A clock is checkable to the minute: 10:10 means the hour hand just past 10, the minute hand on 2.

mistral-large-4-1t-a52b

unjudged

no svg

nemotron-3-nano-omni-30b-a3b-reasoning

unjudged

no svg

duo

Two characters mid-action, not posed: a dog driving a jet ski, a raccoon holding on.

mistral-large-4-1t-a52b

unjudged

no svg

nemotron-3-nano-omni-30b-a3b-reasoning

17/50

monalisa

Leonardo's portrait: three-quarter bust, folded hands, winding river landscape behind.

mistral-large-4-1t-a52b

unjudged

no svg

nemotron-3-nano-omni-30b-a3b-reasoning

14/50

pelican

Everyone knows what a pelican looks like: large beak with throat pouch, long neck, webbed feet.

mistral-large-4-1t-a52b

unjudged

no svg

nemotron-3-nano-omni-30b-a3b-reasoning

17/50