Archive — US — techcrunch
Anthropic releases Opus 5.5 with lower prices and Fable-level performance
Anthropic released Opus 5.5, a cutting-edge AI model that outperforms its predecessor and Fable, priced lower at $20 per million tokens. This launch reflects CEO Dario Amodei's commitment to pacing AI advancements for safety. Opus 5.5 introduces better clarity in communication while maintaining robust safeguards against misuse in sensitive areas like biology and cybersecurity.
Key points
- Opus 5.5 launched as Anthropic's top-tier model with improved performance and lower costs.
- Price reduced to $20 per million tokens, down from $25 for the previous model.
- Faster processing speed and reduced computational demands compared to earlier versions.
- Significant enhancements in communication style; less jargon and clearer message structuring.
- Launch follows Opus 5's release, with additional tiers (Sonnet 5.5, Haiku 5.5) expected soon.
- Maintains safeguards similar to Fable, preventing misuse in biology and cybersecurity sectors.
- CEO emphasizes a measured approach to AI advancements to allow risk prevention to catch up.
Why it matters
- Reflects ongoing trends in AI, emphasizing transparency, safety, and responsible deployment.
- Aligns with 2025 IT trends focused on security and ethical considerations in AI development.
- Follows a growing movement within tech to balance innovation with regulatory and safety measures.
- Highlights industry responses to calls for responsible AI following recent global concerns over AI risks.
The launch of Opus 5.5 underscores Anthropic's dedication to balancing AI capability advancement with safety measures, setting a precedent for future AI developments amidst rising concerns over ethical applications.