Typst — Modern Markup-Based Typesetting System
Typst is a new markup-based typesetting system that is powerful and easy to learn. A modern alternative to LaTeX with fast incremental compilation, friendly error messages, and a scripting language built in. Write beautiful papers, theses, and documents without the LaTeX pain.
What it is
Typst is a markup-based typesetting system written in Rust, designed as a modern alternative to LaTeX. Created by Martin Haug and Laurenz Maedje as a master thesis project at TU Berlin, Typst compiles documents in milliseconds rather than the seconds or minutes LaTeX requires. It supports math equations, bibliographies, figures, tables, and custom templates out of the box.
Typst targets students writing theses, researchers preparing papers, and professionals who need polished PDF output without wrestling with LaTeX package conflicts. Its syntax is closer to Markdown than to TeX, making the learning curve gentler for newcomers.
How it saves time or tokens
Typst's incremental compilation watches your source file and regenerates only the parts that changed. A document that takes LaTeX 30 seconds to compile rebuilds in under 100 milliseconds with Typst. Error messages point to the exact line and explain what went wrong in plain English, eliminating the time spent deciphering cryptic LaTeX logs. The built-in scripting language handles loops, conditionals, and data-driven content without external toolchains.
How to use
- Install Typst via your package manager:
brew install typston macOS,cargo install --locked typst-clivia Rust, orscoop install typston Windows. - Create a
.typfile with your document content using Typst markup syntax. - Compile to PDF with
typst compile paper.typor usetypst watch paper.typfor live reload during editing.
Example
#set page(paper: "a4", margin: 2cm)
#set text(font: "New Computer Modern", size: 12pt)
#set heading(numbering: "1.1")
= Introduction
This paper explores how *Typst* can replace LaTeX.
Compilation is _instant_ -- no more waiting.
== Background
$ E = m c^2 $
#figure(
table(
columns: 3,
[*Tool*], [*Language*], [*Speed*],
[Typst], [Rust], [< 100ms],
[LaTeX], [TeX], [10-60s],
),
caption: [Compilation speed comparison],
)
#bibliography("refs.bib")
Related on TokRepo
- AI Tools for Documentation -- discover AI-powered documentation workflows that pair well with Typst output
- AI Tools for Coding -- explore coding assistants that can generate Typst templates programmatically
Common pitfalls
- Typst's package ecosystem is smaller than LaTeX's CTAN repository, so niche academic journal templates may not exist yet.
- Some LaTeX features like TikZ diagrams have no direct Typst equivalent; you may need to export SVGs from other tools.
- Typst files are not directly accepted by most academic publishers, so you will still need to export to PDF for submission.
Frequently Asked Questions
Typst compiles in milliseconds instead of seconds, provides clear error messages, and uses simpler syntax. However, LaTeX has a larger package ecosystem and wider publisher acceptance. For most academic writing, Typst handles math, citations, and figures just as well.
Yes. Typst supports BibTeX and Hayagriva bibliography files natively. You add a bibliography command at the end of your document and use citation keys inline, similar to LaTeX.
No. Typst has its own template format and cannot directly use .sty or .cls files from LaTeX. You need to recreate templates in Typst syntax, though the process is simpler than writing LaTeX packages.
Typst outputs PDF by default and also supports SVG. It does not produce HTML or DOCX natively, though community tools exist for conversion.
Yes. The Typst compiler is open source under the Apache-2.0 license. There is also a web-based collaborative editor at typst.app that offers a free tier.
Citations (3)
- Typst GitHub— Typst is a modern typesetting system started as a master thesis at TU Berlin
- Typst Documentation— Typst supports incremental compilation and scripting
- Typst Blog— Typst uses X25519-based encryption for its web service
Related on TokRepo
Discussion
Related Assets
Kornia — Differentiable Computer Vision Library for PyTorch
Kornia is a differentiable computer vision library built on PyTorch that provides GPU-accelerated implementations of classical vision algorithms including geometric transforms, color conversions, filtering, feature detection, and augmentations, all with full autograd support for end-to-end learning.
AlphaFold — AI-Powered 3D Protein Structure Prediction
AlphaFold by Google DeepMind predicts three-dimensional protein structures from amino acid sequences with atomic-level accuracy, enabling breakthroughs in drug discovery, enzyme engineering, and structural biology research.
Flash Attention — Fast Memory-Efficient Exact Attention for Transformers
Flash Attention is a CUDA kernel library that computes exact scaled dot-product attention 2-4x faster and with up to 20x less memory than standard implementations by using IO-aware tiling to minimize GPU memory reads and writes.