VIBE
← Back to Leaderboard
AI AgentsAGENT
AI AgentsOpen SourceAGENT3h ago253

About

ARCgentica is an agentic AI system that solves ARC-AGI-2 challenges by deploying LLM-powered sub-agents to analyze input-output grid examples, write Python transform programs, and evaluate them against test inputs. It achieved 85.28% on the ARC-AGI-2 public evaluation using Claude Opus 4.6, making it one of the top-performing open solutions to this benchmark. It's useful for researchers and developers studying AI reasoning, program synthesis, and abstract pattern recognition.

Why it made the leaderboard

If you're researching symbolic AI, program synthesis, or multi-agent LLM architectures, ARCgentica gives you a fully open-source, auditable implementation that achieves one of the highest publicly reported scores on ARC-AGI-2 — a benchmark specifically designed to resist LLM shortcuts. The sub-agent architecture that writes and evaluates Python transformation programs is a concrete, reusable pattern for anyone building reasoning-heavy agent systems.

Tags

arc-agillmagentic-aiprogram-synthesisbenchmarkingpythonclaudeabstract-reasoning

Tech Stack

Python

Comments

No comments yet.