AnChain.AI reported that ARC-AGI-3 reached 99.9% in its development.
Public source
Publisher name
Public post
“Welcome to the AGI era.” - Greg at OpenAI today’s #astra launch announcement I’ve seen the significant upgrade in our development esp in codex. The way we build softwar…
Company
AnChain.AI
Blockchain forensics and crypto tracing experts for fraud, AML & litigation supporting law firms, accountants, victims.
- Industry
- Computer and Network Security
- Location
- San Jose, US
- Company size
- 11–50 employees
About AnChain.AI
AnChain.AI (HQ in San Francisco) is an award winning Agentic AI company specializing in fraud investigation and AML compliance. Trusted by global regulators, law enforcement, and financial institutions including the U.S. SEC, IRS, and FinCEN, AnChain.AI’s software has underpinned over $870 million in seizures and recoveries worldwide. Founded in 2018 by enterprise SaaS veterans, AnChain.AI is backed by top venture capital firms from both Silicon Valley and Wall Street, and was recognized as a 2023 RSAC Innovation Sandbox Finalist, Top 10 Fintech Innovation Lab startup NYC. AnChain.AI is building superintelligence to fight fraud.
See moreLatest activity
Latest activity from AnChain.AI
6 signals
Research & Knowledge
AnChain.AI reported that Terminal-Bench 4.0 reached 57.9% in its development.
Research & Knowledge
AnChain.AI reported that AutomationBench reached 41.4% in its development, more than 2× GPT-5.6 Sol.
Research & Knowledge
AnChain.AI reported that OSWorld 2.0 computer use reached 72.6% in its development.
Discover more
Similar signals
Similar public activity from other companies.
Research & Knowledge
KuppingerCole Analysts
KuppingerCole Analysts reports that ARC-AGI-3 achieved a benchmark of 98.6 to 99.9 across the AGI era.
Research & Knowledge
Web3 Foundation
Web3 Foundation published a feasibility study on arXiv involving an agentic AI system simulated mobility traces anchored around participants' real home and work locations, identifying 18 out of 43 participants overall.
Research & Knowledge
Cequence Security
Cequence Security published a research claim that a global financial services firm ran their first AI discovery scan expecting a manageable number of agents, finding 740% more MCP servers, 800% more AI agents, and 600% more LLM providers than anticipated.
Research & Knowledge
Ridge Security Technology Inc.
Ridge Security Technology Inc. published Ken Huang's new analysis breaking down the full methodology, cost-per-run, token efficiency, and sovereignty implications in his new analysis built on benchmark data.
Research & Knowledge
Cybersecurity Insiders