ARC Prize announces ARC-AGI-2 and ARC Prize 2025

Announced ARC-AGI-2 and ARC Prize 2025 — a human panel averaged 60%, while o3-preview-low scored 4%, r1 0.3% and pure LLMs 0.0%

Why it matters

Made score and cost joint leaderboard metrics

Source — Primary sources first; secondary reports are labelled “Press”

Officialarcprize.orghttps://arcprize.org/blog/announcing-arc-agi-2-and-arc-prize-2025

Added in Week 35, 2026 · See that week

Same week

xAI acquires X Corp in an all-stock dealPressOn the Biology of a Large Language ModelOfficialOpenAI adopts MCP (Agents SDK)PressMCP releases spec 2025-03-26Official

Nearby on the same axis

Controversy over unreleased experimental Llama 4 Maverick behind Arena scorePressCohere Labs and others publish The Leaderboard IllusionPaperEpoch AI posts a clarification on how FrontierMath was made and is usedOfficialTechCrunch reports criticism of Epoch AI for not disclosing OpenAI fundingPress