The March 2026 update of the Python extension for Visual Studio Code introduces two significant enhancements: the ability to search for symbols inside installed packages and an experimental Rust-based parallel indexer. These features aim to streamline code navigation and boost IntelliSense performance, especially for large projects. Below, we answer the most common questions about what's new and how you can leverage these improvements.
What new feature helps you find symbols in installed packages?
With the March 2026 release, Pylance can now include symbols from packages installed in your active virtual environment when you search using Workspace Symbol search (shortcut Ctrl+T or Cmd+T). This means you can quickly locate where a function or class is defined—even if it resides in a third-party library—without leaving VS Code or consulting external documentation. For libraries that lack a py.typed file, only symbols exported via __init__.py or __all__ are included, keeping results relevant. This feature is opt-in by default to preserve performance in everyday use.
How do I enable symbol search for installed packages?
To activate this feature, navigate to VS Code settings (Ctrl+, or Cmd+,), search for “Include Venv In Workspace Symbols,” and check the box under Python › Analysis. Once enabled, your Workspace Symbol search will surface symbols from all packages in the active virtual environment’s site-packages directory. You can further fine-tune indexing depth per package using the setting Python › Analysis: Package Index Depths, which controls how deeply Pylance probes into sub-modules. This gives you tailored control over code exploration while keeping the default experience unchanged for others.
What performance impact does the symbol search feature have?
Because indexing installed packages can affect overall performance, especially on projects with many dependencies, the feature is deliberately opt-in. The new setting Package Index Depths lets you limit the depth of indexing per package, so you can balance rich exploration against resource usage. For most setups, enabling it adds minimal overhead, but it’s recommended for developers who frequently work with unfamiliar libraries or large codebases. If you experience slowdowns, you can disable the feature or reduce indexing depth.
What is the experimental Rust-based parallel indexer?
This release introduces an experimental setting that switches Pylance’s indexer—the engine behind completions, auto-imports, and symbol search—to a Rust-based parallel implementation running out-of-process. In internal testing, this new indexer delivers on average 10× faster performance on large Python projects. The improvement is most noticeable after opening a workspace, providing quicker completions and a more responsive IntelliSense experience overall. Because it’s experimental, you need to enable it explicitly to validate stability and gains across diverse project setups before it becomes the default.
How do I enable the Rust-based parallel indexer?
To try it, open VS Code settings (Ctrl+, or Cmd+,), search for “Parallel Indexing,” and check the box labeled Enable Parallel Indexing (Experimental) under Python › Analysis. Alternatively, add the following to your settings.json:"python.analysis.enableParallelIndexing": true
After enabling, reload VS Code via the Command Palette (Ctrl+Shift+P → Reload Window) to ensure a clean start of the new indexer. This setting has the greatest impact on larger projects; small projects may see little difference.
What should I expect when using the parallel indexer?
Once enabled, the new Rust-based indexer runs in the background and should provide noticeably faster completions and symbol searches, especially on large codebases. The performance boost is most pronounced immediately after opening a workspace, where traditional indexing can be sluggish. However, because it’s still experimental, some edge cases or project configurations might behave differently. The development team encourages users to share feedback on their experience—positive or negative—to help refine the feature before a stable rollout. If you encounter issues, you can easily disable the setting and revert to the default indexer.