← Back to leaderboard

deepseek-r1:14b

10 GB measured VRAM · 8-12GB tier · rag-gauntlet-v1
Exact passages, questions, and expected answers for this category are held back from publication to avoid training-data contamination — see About for why. This page shows per-tier pass/fail counts and timing only.
Tier 1 — floor 5/5
234 tokens · 7.881s (representative seed)
Tier 2 — restraint (refusal) 5/5
334 tokens · 4.131s (representative seed)
Tier 3 — distractor resistance 5/5
270 tokens · 3.405s (representative seed)
Tier 4 — multi-passage synthesis 5/5
644 tokens · 7.908s (representative seed)
Tier 5 — precise extraction 5/5
240 tokens · 3.04s (representative seed)
Tier 6 — long-context needle 5/5
261 tokens · 3.492s (representative seed)