Anthropic just launched Claude Opus 4 and Claude Sonnet 4, their most powerful models so far.
- Opus 4: Best coding model (72.5% SWE-bench), built for long, complex tasks and agents.
- Sonnet 4: Major upgrade to 3.7, with better reasoning and faster, sharper code (72.7%).
New features:
– Tool use during reasoning
– Parallel tools + memory
– VS Code & JetBrains integrations
– New API tools (code exec, MCP, Files, caching)
Same pricing. Available via API, Bedrock, and Vertex AI.
Full details: anthropic.com/news/claude-4
What’s your take on it?