Fetching from the wire…
Public story · 2026-02-22 · source-backed
The biggest single-benchmark jump in a frontier model update: 37.6% → 68.8% on ARC-AGI-2. For comparison, GPT-5.2 scored 54.2%, Gemini 3 Pro 45.1%. ARC-AGI-2 measures novel problem-solving on adversarially constructed tasks. A near-doubling suggests genuine reasoning improvement, not benchmark optimization. ARC Prize
Each link below shares sources, entities, or timing with this story.
Gemini partners with NotebookLM / Shared entities / Shared topic / Earlier coverage
Linked by a graph relationship (Gemini partners with NotebookLM); both cover AGI, ARC, Claude Opus, Gemini; overlapping topics (arc-agi-2, benchmark, claude, gemini).
Claude Opus built by Anthropic / Shared entities / Shared topic / What happened next
Linked by a graph relationship (Claude Opus built by Anthropic); both cover AGI, ARC, ARC Prize, GPT; overlapping topics (benchmark, frontier, gpt-5, model).
Claude Code benchmarked against Gemini / Shared entities / Shared topic / What happened next
Linked by a graph relationship (Claude Code benchmarked against Gemini); both cover AGI, ARC, Claude Opus, GPT; overlapping topics (benchmark, claude, model).
Codex competes with Gemini / Shared entities / Shared topic / What happened next
Linked by a graph relationship (Codex competes with Gemini); both cover AGI, ARC, Claude Opus, Gemini; overlapping topics (gemini, model).
Linked by a graph relationship (Codex competes with Gemini); both cover AGI, ARC, GPT; overlapping topics (claude, frontier, model).
Claude Code benchmarked against Gemini / Shared entities / Shared topic / What happened next
Linked by a graph relationship (Claude Code benchmarked against Gemini); both cover AGI, ARC, Claude Opus; overlapping topics (benchmark, claude, model).
Claude Code benchmarked against Gemini / Shared entities / Shared topic / Earlier coverage
Linked by a graph relationship (Claude Code benchmarked against Gemini); both cover Claude Opus, Gemini, GPT; overlapping topics (benchmark, claude, model).
Gemini built by Google / Shared entities / Shared topic / What happened next
Linked by a graph relationship (Gemini built by Google); both cover AGI, ARC, Gemini; overlapping topics (arc-agi-2, model).