Apr 20, 2026
Anthropic's newly released Claude Opus 4.7 uses a new tokenizer that breaks text into up to 47% more tokens than its predecessor, resulting in significantly higher costs per request despite maintaining the same per-token pricing as Opus 4.6.
Anthropic launched Claude Opus 4.7, its most powerful generally available model, featuring significant improvements in software engineering tasks while deliberately reducing certain cybersecurity capabilities during training.
Anthropic restricted access to its Claude Mythos AI model after it identified thousands of zero-day vulnerabilities across major operating systems, browsers, and cryptography libraries that DeFi infrastructure depends on. The company warned AI models have reached coding capability surpassing most skilled humans at finding exploits.