**Anthropic is developing advanced reasoning models** (Claude Sonnet and Claude Opus versions) that can “think” harder and longer than existing reasoning AI, with releases expected in the coming weeks[4]. This represents a significant escalation in the competition between major AI labs to create more sophisticated reasoning capabilities[4].

References:
1. https://www.geekwire.com
2. https://www.technewsworld.com
3. https://techcrunch.com
4. https://www.theinformation.com/technology