Windsurf + Claude 3.7 Sonnet (Thinking) is 🔥 Watch how it created a website with one simple prompt. An agentic coder combined with a top reasoning model is as good as it gets. Jump to 1:40 for the result.

I've been impressed by how good Windsurf agentic capabilities are. I believe it's the best out there that I've used. It's the first time I've felt that a coding agent can create useful working code. The predictive capabilities inside the editor are also rapidly improving.

I also noticed Claude 3.7 Sonnet is incredible at generating mermaid diagrams. I love this sequence diagram it generated for a tutorial on function calling that I am preparing. I literally just prompt: "Generate a mermaid sequence diagram for the flow in x script" in Windsurf.

Been testing Claude 3.7 Sonnet (thinking) the whole day and now I don’t think I can switch back to Claude 3.5. If you switch between the models you will note a huge difference in code quality. Anthropic really are serious on their model being the best at code. It shows!

Boost, surf, and weave your way through The Art of Flight, an arcade game about flying multiple ships at the same time. With solo, local co-op, and a leaderboard, there are tons of ways to play. Wishlist on Steam today!

Just about any AI coding IDE / platform can create amazing websites with one single prompt now with Claude 3.7 Sonnet. Getting similar amazing results with @Replit

That’s impressive! Combining a powerful coding agent with top-tier reasoning like Claude 3.7 Sonnet makes building things so much easier.

Waoooo…..

I haven't reviewed Sonnet-3.7's coding benchmark numbers, but I suspect it's less capable than Sonnet-3.5. I tested Sonnet-3.7 on a simple task, and the results were subpar and somewhat outdated. What's your opinion?

The figure is what they reported. So far in my experiments, I feel like this model is superior. I will continue to experiment some more and see where the model potentially struggles.

that’s some next-level wizardry, like coding with magic!

Amazing!