Inkling Small Lands Within A Point Of Inkling On The Artificial Analysis Intelligence Index With Less Than A Third Of The Parameters
Source
artificialanalysis.ai
Author
Artificial Analysis articles
Date
Why it matters
Inkling Small scores 40 with 276B total (12B active) parameters and about 24K output tokenThe chunk of text a model reads and writes in — roughly three-quarters of a word — and the unit AI usage is billed in.Full definition → per task, a small open-weights option near its larger sibling.
Key takeaways · AI-distilled
Artificial Analysis reports that Inkling Small nearly matches Inkling's overall index score with fewer parameters, but loses ground on banking tasks and factual knowledge. The aggregate score hides these task differences.
Inkling Small's stronger AA-Briefcase Elo came largely from presentation. Its rubric pass rate was 20% versus Inkling's 19%, while it used 34 rather than 81 turns per task on average.
Inkling Small's weaker AA-Omniscience score reflects fewer correct answers, despite a slightly lower hallucinationWhen a model states something false with full confidence — inventing facts, citations, or APIs that don't exist.Full definition → rate than Inkling. A lower hallucination rate therefore does not imply a stronger factual-knowledge score.
Terms in this piece · Glossary
mixture-of-experts — A model built from many specialist sub-networks where only a few activate per token, giving big-model capability at small-model running cost.
token — The chunk of text a model reads and writes in — roughly three-quarters of a word — and the unit AI usage is billed in.
hallucination — When a model states something false with full confidence — inventing facts, citations, or APIs that don't exist.
