Is Sonnet 4 the new default for devs?
Matt V
May 26, 2025
ANTHROPIC
Anthropic has launched Claude Opus 4 and Claude Sonnet 4, two upgraded AI models designed to push the boundaries of coding, reasoning and tool use.
Whether you’re building apps, debugging code, or running complex workflows, these models are built to keep up.
Claude Opus 4 is heavy, designed for long, complex tasks and agent-style workflows.
Developers say it performs better on big coding challenges than anything else on the market.
It’s already been successfully tested on long-running projects, working for hours without dropping the ball.
Claude Sonnet 4 is faster and more efficient, making it great for day-to-day coding and writing tasks.
It’s a direct upgrade from Sonnet 3.7 and now powers tools like GitHub Copilot, offering improved instruction-following and more refined output.
Both models come with:
Tool use during extended tasks (like web searches)
Parallel tool use and better memory for long-term work
More accurate handling of complex instructions
Tool time, baby
On top of that, Claude Code is now available to all.
It integrates with VS Code and JetBrains, letting you review, edit, and debug code right inside your editor.
You can even tag Claude on GitHub pull requests to fix CI errors or respond to reviews.
There’s a new SDK too, so you can build your own Claude-powered coding tools.
Pricing stays the same: Opus 4 at $15/$75 per million tokens and Sonnet 4 at $3/$15. Both are available through Anthropic’s API, Amazon Bedrock, and Google Cloud’s Vertex AI.
Sonnet 4 is for when you want speed but still need brains.