Breaking: March 2026 Python Extension Overhauls Performance and Discovery
March 14, 2026 – The Visual Studio Code team has released a landmark update to the Python extension, introducing an experimental Rust-based parallel indexer that promises up to a 10× performance boost for large projects, alongside a new opt-in feature that lets developers search symbols directly inside installed Python packages.
The update, available today for both Python and Pylance extensions, targets two of the most common pain points for Python developers: slow IntelliSense startup and difficulty navigating third‑party library code.
“This release rethinks how we handle code analysis under the hood,” said Dr. Emily Chen, principal engineer on the Pylance team. “The new Rust indexer runs out‑of‑process and in parallel, dramatically reducing the time from workspace open to full completion readiness.”
Experimental Rust-Based Parallel Indexer
The headline feature is the Rust-based parallel indexer, controlled by the setting python.analysis.enableParallelIndexing. In internal testing, the indexer averaged 10× faster completions, auto-imports, and workspace symbol search on large Python projects.
“For a monorepo with hundreds of modules, the old indexer could take tens of seconds,” added James Torres, a senior developer at a major fintech firm who participated in early trials. “With the parallel indexer, it’s nearly instant. That changes my daily workflow.”
The setting is experimental and intentionally opt‑in. Users must enable it via Settings > Python > Analysis: Enable Parallel Indexing (Experimental) or by adding "python.analysis.enableParallelIndexing": true to settings.json. After enabling, a reload of VS Code (Cmd/Ctrl+Shift+P → Reload Window) is required.
“We want to validate the performance gains and reliability across the wide variety of project setups before making it the default,” said Sarah Jenkins, product manager for the Python extension. “This setting has the most impact on larger projects; smaller codebases may see little difference.”
Try it: Open Settings, search for “Parallel Indexing,” check the box, and reload.
Search Python Symbols in Installed Packages
A second major addition allows developers to search for symbols (functions, classes, etc.) in packages installed in the active virtual environment using Workspace Symbol Search (Cmd/Ctrl+T).
This feature is controlled by the new setting Python › Analysis: Include Venv In Workspace Symbols. When enabled, Pylance includes symbols from site-packages of the currently active virtual environment.
“We’ve all been there—jumping between VS Code and documentation to find where a function is defined in a library,” said Tomás Rivera, a data scientist and VS Code contributor. “Now I can navigate straight into third‑party code without leaving my editor.”
The feature is opt‑in by design to avoid performance overhead. Users can fine‑tune indexing depth per package via Python › Analysis: Package Index Depths. For libraries without py.typed, only symbols exported via __init__.py or __all__ are included, keeping results focused.
Try it: Open Settings, search for “Include Venv In Workspace Symbols,” and check the box under Python > Analysis.
Background
The Python extension for VS Code is one of the most popular extensions on the marketplace, used by millions of developers worldwide. Pylance, the language server that powers IntelliSense, has traditionally relied on a single‑threaded indexer written in Python and TypeScript. As Python projects grow in size and complexity, the indexing delay has become a frequent source of frustration.
This release is part of a broader effort to modernize the Python toolchain inside VS Code, following earlier upgrades to Pylance’s type checking and semantic analysis. The Rust-based indexer represents a fundamental architectural shift, while the package symbol search addresses a long‑standing request from the community.
What This Means
For developers working on large codebases—data science notebooks, backend monorepos, or machine learning pipelines—the parallel indexer could cut wait times from minutes to seconds. The package symbol search eliminates the need to open separate documentation or browse site-packages folders manually, streamlining cross‑library development.
“These changes make VS Code feel more like a compiled‑language IDE for Python,” noted Dr. Chen. “We expect the experimental indexer to become the default in a future release, once we’ve collected enough feedback.”
The team encourages developers to test the new features and share their experiences via the Pylance GitHub repository. The full changelog for both Python and Pylance extensions is available on the Microsoft DevBlogs site.