Dimension scores are derived from public data and fields; weighted into the composite. Reference only.
LFortran is a modern open-source Fortran compiler based on LLVM and released under the BSD license. It is positioned not only as a traditional offline compiler, but also as a tool with strong interactive execution capabilities, aiming to make Fortran usable for exploratory computing in a way similar to Python, Julia, or MATLAB. The official documentation clearly states that the project is still in the alpha stage and is not yet expected to handle third-party code reliably, so users are encouraged to actively participate in bug reporting and fixes.
In terms of functionality, LFortran can already parse the full Fortran 2018 syntax into an AST and supports formatting source code back into Fortran. It provides a command-line REPL and a Jupyter kernel, making it suitable for researchers who want to quickly experiment with algorithms. On the compilation side, it can create executable files and offers multiple backends, including LLVM, C++, x86, and WebAssembly. The C++ backend can generate readable C++, while the WebAssembly backend supports in-browser testing. The project also emphasizes a modular AST and ASR design, allowing it to be reused as a library by other development tools.
LFortran supports Linux, Mac, Windows, and WebAssembly, with these platforms regularly tested through CI. Community support channels include Zulip, mailing lists, and Fortran Discourse. The official criterion for entering beta is whether it can reliably compile 10 third-party production codebases. The main text currently shows progress at 9/10 and lists work on Minpack, fastGPT, dftatom, parts of SciPy, stdlib, SNAP, PRIMA, POT3D, LAPACK, fpm, and others, indicating a relatively high level of transparency.
The main text does not mention commercial pricing, paid plans, or enterprise support. As a BSD licensed open-source project, its main value lies in being free to use, auditable, and open to secondary development, making it suitable for universities, research institutions, and toolchain developers.
Its strengths include a rare interactive Fortran experience, Jupyter support that is friendly to scientific computing, and a multi-backend, modular architecture that is useful for compiler research. Its weaknesses are also clear: it remains in alpha, full Fortran 2018 compilation support is still under development, and there is limited information on production-code compatibility and enterprise support. It is better suited to Fortran/HPC researchers, early adopters, and developer-tool authors than to teams seeking a stable production compiler toolchain. The latter may still want to keep an eye on GFortran or Flang.
The main text does not provide information about mainland China network access, mirrors, or payment options. Since it depends on the official website, GitHub, and ecosystems such as Jupyter/conda, the actual access experience may be affected by local network conditions. However, this cannot be determined from the text alone, so it is recommended to test the download and build process locally.
⚠ This review is compiled from public sources and does not constitute a purchase recommendation. Verify all facts on the vendor's official site. Verify on lfortran.org official site.
lfortran.org is an United States Dev Tools provider. TG4G tracks its product information, an overall rating of 8.0/10, and a China-accessibility score of China direct-connect friendly. Click "Visit Official Site" to reach lfortran.org directly.