Available Models
Claude Sonnet 4.5 Anthropic’s most intelligent Sonnet model. Hybrid reasoning with fast and extended thinking modes. Excels at real-world coding, agents, and computer use with 200K context.
Claude Opus 4.5 Anthropic’s most capable model for complex software engineering and agentic workflows. 80.9% on SWE-bench Verified. Optimized for long-horizon computer use tasks.
Claude Sonnet 4 Major upgrade to Sonnet 3.7 with 72.7% on SWE-bench Verified. Ideal balance of capability and efficiency for everyday coding and business tasks.
Claude Haiku 4.5 Anthropic’s most efficient model matching Claude Sonnet 4 performance. >73% on SWE-bench Verified. Optimized for high-throughput coding and computer use tasks.
Claude 3.5 Haiku Fast and capable model surpassing Claude 3 Opus on most benchmarks. 40.6% on SWE-bench Verified. Best value for coding assistance and quick tasks.
Claude 3 Haiku Anthropic’s fastest model for near-instant responses. Ideal for lightweight tasks, customer service, and content moderation at the lowest cost in the Claude family.