Type checkers · 5 tools
Top 5 type checkers for gradually adding types to dynamic-language codebases
This shortlist starts with mypy for Python and Sorbet for Ruby, then covers Pyright, Flow and TypeScript. It is for teams working in Python, Ruby and JavaScript that want to add static types to existing code in stages instead of rewriting it. Each entry shows what the vendor documents about incremental adoption and where that documentation stops.
| # | Tool | Best for | Key feature | Platforms / deployment | Pricing | More |
|---|---|---|---|---|---|---|
| 1 | mypyPython Project (mypy) | Python teams migrating existing code to static types step by step | Code can be migrated incrementally, mixing static and dynamic typing | Any Python VM | See vendor pricing page | Introduction |
| 2 | SorbetStripe | Ruby teams adding types file by file | Adoption one file at a time | Command line and editors via the Language Server Protocol; easy to add to CI | See vendor pricing page | Introduction |
| 3 | PyrightMicrosoft | Python teams with large codebases who want a fast checker in VS Code | Designed for high performance and large codebases | Command line and Visual Studio Code extension | See vendor pricing page | Introduction |
| 4 | FlowMeta | JavaScript and React teams that want a typed dialect with React-specific syntax | TypeScript-like syntax such as keyof, readonly props and unknown | Can be tried in the browser; other platforms not stated on official pages | See vendor pricing page | Introduction |
| 5 | TypeScriptMicrosoft | JavaScript teams adding types to an existing project | Types can be adopted gradually in an existing JavaScript project | Compiles to JavaScript that runs in browsers, Node.js, Deno and Bun | See vendor pricing page | Introduction |
1 mypy
By Python Project (mypy)
Mypy is an optional static type checker for Python that combines dynamic and static typing.
Why it is here: It sits first because gradual migration is part of its stated design: programs start dynamically typed and gain static typing later.
- Best for
- Python teams migrating existing code to static types step by step
- Key features
- Code can be migrated incrementally, mixing static and dynamic typing
- Checks annotations conforming to PEP 484
- Infers variable types, often needing only function signatures
- Library stubs for checking library usage
- Runs on any Python VM with basically no runtime overhead
- Limits
- Documented for Python only; teams working in Ruby or JavaScript need a different tool. Editor integration details are not described on the page provided, so check the vendor documentation.
- Platforms / deployment
- Any Python VM
- Pricing
- See vendor pricing page (not quoted here; facts checked 21 November 2026, Asia/Dhaka)
Official website (mypy-lang.org) · Read our introduction to mypy
2 Sorbet
By Stripe
Sorbet is a fast, gradual static type checker for Ruby built at Stripe.
Why it is here: It is the closest fit for Ruby because it works with normal Ruby and allows adoption one file at a time.
- Best for
- Ruby teams adding types file by file
- Key features
- Adoption one file at a time
- Backwards-compatible signature syntax
- Opt-in stricter static checks
- Language Server Protocol editor integration
- sorbet-runtime gem for dynamic runtime checks
- Limits
- Documented for Ruby only. Teams that need Python or JavaScript tooling should look at the other entries and check the vendor documentation.
- Platforms / deployment
- Command line and editors via the Language Server Protocol; easy to add to CI
- Pricing
- See vendor pricing page (not quoted here; facts checked 21 November 2026, Asia/Dhaka)
Official website (sorbet.org) · Read our introduction to Sorbet
3 Pyright
By Microsoft
Pyright is a full-featured, standards-based static type checker for Python from Microsoft, with a command-line tool and a VS Code extension.
Why it is here: It follows mypy because its documentation emphasizes performance on large codebases, while the pages provided say less about incremental adoption itself.
- Best for
- Python teams with large codebases who want a fast checker in VS Code
- Key features
- Designed for high performance and large codebases
- Command-line tool and Visual Studio Code extension
- Browser-based Pyright Playground
- Weekly benchmark against other Python type checkers
- Core type checking issues belong to Pyright, language service issues to Pylance
- Limits
- The pages provided do not describe a gradual-adoption workflow, so teams planning one should check the vendor documentation. Documented for Python only.
- Platforms / deployment
- Command line and Visual Studio Code extension
- Pricing
- See vendor pricing page (not quoted here; facts checked 21 November 2026, Asia/Dhaka)
Official website (github.com/microsoft/pyright) · Read our introduction to Pyright
4 Flow
By Meta
Flow is a typed dialect of JavaScript with a static type system and first-class language constructs, built for safety and scale.
Why it is here: It is the JavaScript entry that fits teams in the React ecosystem, but its documented highlights are language features rather than an incremental migration path.
- Best for
- JavaScript and React teams that want a typed dialect with React-specific syntax
- Key features
- TypeScript-like syntax such as keyof, readonly props and unknown
- Conditional types, mapped types and type guards
- Component and renders built into the syntax for React
- Exhaustively checked pattern matching via match
- Used in production at Meta across millions of files of JavaScript and React
- Limits
- Documented as a typed dialect of JavaScript; the pages provided do not describe file-by-file adoption. Teams planning a staged rollout should check the vendor documentation.
- Platforms / deployment
- Can be tried in the browser; other platforms not stated on official pages
- Pricing
- See vendor pricing page (not quoted here; facts checked 21 November 2026, Asia/Dhaka)
5 TypeScript
By Microsoft
TypeScript is a strongly typed programming language from Microsoft that builds on JavaScript and catches errors in the editor before code runs.
Why it is here: Its documentation covers gradual adoption clearly, but it is a full language that compiles to JavaScript rather than a checker designed only for incremental adoption, so it sits last here.
- Best for
- JavaScript teams adding types to an existing project
- Key features
- Types can be adopted gradually in an existing JavaScript project
- A // @ts-check comment makes the editor show type errors in a JavaScript file
- JSDoc comments can supply type information for plain JavaScript
- Existing JavaScript code is valid TypeScript
- Type inference provides tooling without extra annotations
- Limits
- Documented for JavaScript projects; it compiles to JavaScript, so teams adopt it as a language and build step. Python and Ruby teams need a different tool.
- Platforms / deployment
- Compiles to JavaScript that runs in browsers, Node.js, Deno and Bun
- Pricing
- See vendor pricing page (not quoted here; facts checked 21 November 2026, Asia/Dhaka)
Official website (typescriptlang.org) · Read our introduction to TypeScript
How we chose these tools
We considered fit with the use case of gradually adopting static typing in dynamic-language codebases, the capabilities each vendor documents, and the platform and deployment information on official pages. The most weight went to documented support for mixing typed and untyped code, adopting types file by file, and keeping existing code working. Editor integration and performance on large codebases were secondary criteria.
All facts come from official vendor pages, each with a recorded check date. The order reflects editorial fit for this use case: tools whose documentation makes gradual, incremental adoption central to the design come first, and tools where it is one feature among others follow. The list is built from documentation rather than product trials, and where a page is silent on a point we say so. Pricing is not covered here; see each vendor's pricing page.
Order used in this guide: Ordered by how central gradual, incremental adoption is to each tool's design, closest fit first.
Frequently asked questions
How do I add types to an existing Python codebase gradually?
Mypy documents that programs can start dynamically typed and gain static typing later, and that code can be migrated incrementally by mixing static and dynamic typing. It checks annotations conforming to PEP 484.
Can Sorbet be adopted one file at a time?
Yes. Sorbet's documentation says it works with normal Ruby and allows adoption one file at a time, with developers able to opt into stricter static checks.
Can I try TypeScript in an existing JavaScript project without renaming files?
TypeScript documents that adding a // @ts-check comment to a JavaScript file makes the editor show type errors, and that JSDoc comments can supply type information for plain JavaScript files.
What is the difference between mypy and Pyright?
Both are static type checkers for Python. Mypy's documentation centers on optional, incremental typing, while Pyright's documentation emphasizes high performance on large codebases and ships a command-line tool and a VS Code extension.
Is Flow the same as TypeScript?
No. Flow is a typed dialect of JavaScript from Meta, while TypeScript is a language from Microsoft that builds on JavaScript. Flow's documentation notes it supports TypeScript-like syntax such as keyof, readonly props and unknown.
- mypy: mypy-lang.org, mypy-lang.org/about.html — checked 21 November 2026
- Sorbet: sorbet.org, sorbet.org/docs/overview — checked 21 November 2026
- Pyright: github.com/microsoft/pyright, github.com/about, github.com/features — checked 21 November 2026
- Flow: flow.org — checked 21 November 2026
- TypeScript: typescriptlang.org, typescriptlang.org/why-create-typescript — checked 21 November 2026
Linthelm is independent and not affiliated with the vendors listed. No vendor paid for inclusion or position. Product names belong to their owners. Features and prices change; confirm details on each official website.