Anthropic announced Claude Mythos Preview on April 7, 2026, as its most capable frontier large language model to date, with exceptional performance in software engineering, reasoning, and especially cybersecurity tasks such as zero-day exploitation. Due to these risks, the company limited initial access to vetted partners through Project Glasswing for cybersecurity-focused use only, with private preview availability also on Google Cloud Vertex AI; it has not gone generally available. Subsequent safer models like Claude Opus 4.8 and Fable 5 reached wider release in May and June 2026, while Mythos remains restricted. Traders should monitor Anthropic statements on safeguard progress and any expanded partnerships, as these directly influence timelines for broader Mythos-class availability.