From 0% to 36% on Day 1 of ARC-AGI-3
symbolica.ai
March 27, 2026
1 min read
🔥🔥🔥🔥🔥
49/100
Summary
The Agentica SDK by Symbolica achieves a score of 36.08% on ARC-AGI-3, completing 113 out of 182 levels and 7 out of 25 games. This performance surpasses CoT baselines significantly, with Agentica costing $1,005 compared to Opus 4.6's $8,900 for a mere 0.25% score.
Key Takeaways
- The Agentica SDK by Symbolica achieved an unverified competition score of 36.08% on ARC-AGI-3, passing 113 out of 182 playable levels.
- Agentica completed 7 out of the 25 available games in the ARC-AGI-3 competition.
- Agentica's performance of 36.08% was achieved at a cost of $1,005, significantly lower than Opus 4.6's cost of $8,900 for a score of 0.25%.
- The implementation outperformed CoT baselines, which scored 0.2% (Opus 4.6 Max) and 0.3% (GPT 5.4 High).
Coverage & discussions
Hacker News · symbolica.ai
Community discussions
Hacker News discussion ↗
Historical count: 4 · 4 saved in sample · Historical sample
Awaiting analysis
Related Articles
ARC-AGI Leaderboard
Jul 25, 2026
ARC-AGI-3
Mar 25, 2026
Schema Harness Achieves ~99% on Arc‑AGI‑3 Public
Jul 16, 2026
Atlassian’s Mike Cannon-Brookes: AI agents need a name tag—and a human to be accountable for their actions
Oct 19, 2026
Harness Engineering 101: The LLM is the brain, the harness is the body
Oct 19, 2026