ARC-AGI-3
Interactive reasoning benchmark designed to measure human-like intelligence in AI agents
arcprize.org ↗📍 San Francisco, CA
Verified Data
“Estimated 50k-200k+ visits (High peak interest due to 2026 launch).”
“1,454 teams competed in 2025.”
“2-10 employees (Core team includes 8-10 people).”
“Prize pool growth: 2024 ($600k) -> 2025 ($700k) -> 2026 ($2.0M). 1,400+ teams in 2025.”
Company Profile
Contact
@arcprizeStrategic Analysis
Strategy
Operates as a non-profit foundation focused on advancing AGI research through open benchmarks. Leverages major AI lab partnerships (OpenAI, Anthropic, Google) for funding and credibility. Uses competitive prize structure to incentivize breakthrough research in AI reasoning capabilities.
Tactics
Runs annual competitions with escalating prize pools ($600K to $2M) to drive participation. Partners with Kaggle for competition platform infrastructure. Builds community through launch events featuring industry leaders like Sam Altman and François Chollet.
Competitive Positioning
Positions as the definitive benchmark for testing true reasoning vs. pattern matching in AI systems. Differentiates from other AI benchmarks by focusing specifically on novel problem-solving rather than knowledge retrieval or training data memorization.
Marketing Approach
Community-driven growth through AI research networks and academic institutions. High-profile launch events and partnerships with major AI labs for credibility. Uses social media and industry conferences to build awareness among target researchers.
Notable
Founded by François Chollet (creator of Keras), supported by major AI labs including OpenAI and Anthropic
🔗 Source ↗Tech Stack
Next.js
React
GitHubRecent News
Related AI Benchmarking Companies
Discovery Sources
Signals
Evidence
100% score means AI agents can beat every game as efficiently as humans.
$150K - $250K•US / Remote (US)