Claude 3.7 Sonnet's extended thinking rivals Copilot's "Think Deeper"

Anthropic comes at Microsoft Copilot’s “Think Deeper” feature with its own “extended thinking” mode on Claude 3.7 Sonnet.

The Amazon-backed company, which was formed by ex-OpenAI, has recently launched the latest model. It is now available on platforms like Amazon Bedrock and Google’s Vertex AI.

Claude 3.7 Sonnet comes with two modes: a standard mode for quick responses and an extended thinking mode for deeper analysis. It runs almost the same as Copilot’s “Think Deeper” feature.

You can even control how much time the model spends thinking by adjusting token limits. And, it lets you balance speed and quality depending on the task.

“Extended thinking gives Claude 3.7 Sonnet enhanced reasoning capabilities for complex tasks, while also providing transparency into its step-by-step thought process before it delivers its final answer,” Anthropic says.

It does so by taking extra time to break down the problem step by step. The model creates “thinking” blocks, where it shows its internal reasoning before giving a final answer. This helps it think through complex tasks more thoroughly.

But, in multi-turn conversations, only the latest reasoning steps are visible to the model and counted as input tokens to make it more efficient.

Copilot’s Think Deeper, on the other hand, is powered by OpenAI’s reasoning-excelling model, the o1. ChatGPT also has a similar feature with the “Reason” button on the chatbox, and it also has an AI agent called “deep research” which lets you deep dive into complex topics on the internet—not to be confused with the reasoning.  

Other than that, Anthropic introduced Claude Code, a command-line tool that lets developers work directly with the AI. You can do so by delegating tasks like code editing, testing, and GitHub integration.

Claude 3.7 Sonnet is now available on all Claude plans. It includes Free, Pro, and Enterprise, with extended thinking available to all users except on the free tier.