Today, Anthropic announced their latest large language model: Claude 3.5 Sonnet. This model’s reasoning skills outperform Anthropic’s previous flagship model—Claude 3 Opus—while being roughly twice as fast.

Cody, our AI coding assistant, is built on the principle of choice, and we’re committed to bringing the best models and best context into your IDE with Cody. To that end, we’re making Claude 3.5 Sonnet available today in Cody Pro. Pro users can now hot-swap their model to 3.5 Sonnet for chat, commands, and code edits with unlimited usage.

See Claude 3.5 Sonnet in action for coding tasks

Anthropic’s announcement blog cites impressive benchmarks for graduate-level reasoning and coding performance for Claude 3.5 Sonnet, even beating out Claude 3 Opus in both cases. Using s0.dev, we can see how 3.5 Sonnet compares against other high-intelligence models for coding questions using public repositories:

You can try Claude 3.5 Sonnet for yourself on s0.dev to compare it against other LLMs like Claude 3 Opus, Gemini 1.5 Pro, and GPT-4o.

Get started with Claude 3.5 Sonnet and Cody

Claude 3.5 Sonnet is available today for Cody Pro, and you can get started with it by downloading the latest VS Code extension release (v1.22.4). It will also be available in the next JetBrains plugin release in the coming days.

Claude 3.5 Sonnet will also come to Cody Enterprise users in the next Sourcegraph release planned for July 10.

If you don’t yet have Cody, you can get started for free. If you want unlimited access to the latest models—including Claude 3.5 Sonnet—you can upgrade to Cody Pro for $9/month.