I benchmarked what frontier models actually know about 2026 — most of it, they don't

Automated excerpt

Every LLM leaderboard tells you how models score on knowledge from their training data. training cutoff? Gemini 3. 7 Flash leads at 2025-2026 facts. Multi-hop variant: questions whose answer requires joining TWO 2025-2026 pages. retrieval uplift per model (preliminary local data on a 3B pipeline: 0% → 33. 3%).

Selected automatically from source text; not independently written or fact-checked. Read the original for full context.