Tamarin prover

Tamarin Prover team wins the Levchin Prize for Real World Cryptography

Mon, 09 Mar 2026 09:00:00 +0000

We are proud to share that the Tamarin Prover team won the International Association for Cryptologic Research (IACR) Real World Cryptography 2026 Levchin Prize for “developing the Tamarin prover, and its use in the analysis of real world security protocols”.

The award ceremony took place at the IACR Real World Crypto (RWC) conference in Taipei, Taiwan. Watch the award ceremony and the keynote on Tamarin given by Cas Cremers.

We are extremely grateful to all of our developers, contributors, and users, who have made Tamarin a success throughout the years. We look forward to many exciting new features and use cases in the future.

Thank you all!

Tamarin 1.12 released

Sun, 08 Mar 2026 09:00:00 +0000

Tamarin 1.12 is now available with many quality of life features and GUI improvements, including reloading externally edited files, and new zoomable graphs, which can be popped out into a separate window, e.g., for multi-monitor use.

Graph improvements

Graphs are now in a zoomable format (SVG)
There are different levels of abstraction for better readability when zooming
The graphs can be popped out into their own window, for example for full screen or multi-display setups

Other GUI improvements

Lemmas can now be edited inside the GUI
Files can be reloaded from the GUI without having to restart Tamarin
Better well-formedness messages for fact usage, locks, etc.

Additional new features

Improvements to the Tamarin Manual
Macros now also apply to lemmas and restrictions

Backend improvements

Numerous improvements of the regression tests, including:
- automated testing whether Tamarin can parse its own output
- explicit testing for derivation checks which are now disabled for other tests
- replaced oracles with tactics for regression tests
Tamarin now allows Maude up to version 3.5.1
Lots of refactoring
Numerous bugfixes, in particular concerning the parser and pretty printing.
Updated to stack LTS resolver 22.44

`batch-tamarin` is now available

Thu, 30 Oct 2025 09:00:00 +0000

batch-tamarin is a Python wrapper for Tamarin Prover that enables batch execution of protocol verification tasks with reproducibility features. Available now on PyPI. Blog post below!

Context

If you’ve ever worked with Tamarin Prover, you know the pain of running large experiments. Whether you’re comparing different prover versions, testing multiple models, or trying to reproduce someone else’s results, you’ve probably ended up writing custom shell scripts that are hard to maintain and even harder to share.

Researchers have tried to solve this problem in different ways. Some tools like UT-Tamarin help with individual model development, while others like ANSSI’s parser focus on specific analysis workflows. Many research groups even built their own Python wrappers for their papers. But these solutions are usually limited in scope—you might get good lemma-level testing, or nice parsing, or basic orchestration, but rarely all together in a way that works across different experiments.

That’s exactly why batch-tamarin has been built. After observing researchers struggle with the same reproducibility issues over and over, it became clear that a unified solution was needed that could handle real-world research workflows: multiple models, different Tamarin versions, proper resource management, and results that are actually comparable. Instead of yet another paper-specific script, batch-tamarin aims to be the tool that works for any Tamarin experiment, making it easier to focus on the protocol analysis.

Features

batch-tamarin provides a comprehensive solution for reproducible Tamarin workflow orchestration with the following key features:

Declarative Configuration: Define experiments using JSON “recipes” that specify models, Tamarin versions, resources, and options with inheritance support for clean, maintainable configurations
Multi-Version Management: Seamlessly work with multiple Tamarin Prover versions (1.4.0 through 1.10.0)
Intelligent Scheduling: Choose from three scheduling policies (FIFO, SJF, LJF) to optimize resource usage and execution time based on your experimental needs
Per-Lemma Analysis: Automatic parsing of SPTHY models enables fine-grained execution and reporting at the lemma level, with detailed telemetry for each proof attempt
Interactive Setup: Use the init command to generate configuration files through guided prompts, making it easy to get started without manual JSON editing
Semantic Caching: Avoid redundant computations with intelligent caching that considers model content, prover version, and execution parameters
Comprehensive Reporting: Generate detailed reports in HTML, Markdown, LaTeX, or Typst formats with execution timelines, resource usage charts, and automatically rendered attack traces
Validation Tools: Quickly verify model well-formedness and configuration correctness using the check command before launching long-running experiments
Rerun Support: Automatically generate rerun recipes for failed lemmas, allowing easy parameter adjustment without reconfiguring the entire experiment
Cross-Platform Support: Works on macOS and Linux

Technical Details

A full report of the first version of batch-tamarin, including its architecture, design decisions, and implementation details is available in the GitHub repository report, in addition to the documentation markdown in the root of the repository: ARCHITECTURE.md.

Reproducibility ?

Yes, reproducibility was one of the main design goals for batch-tamarin. The tool provides Dockerfiles that can be used to build Docker images with all dependencies needed to run batch-tamarin with a preconfigured environment containing Tamarin Prover. Currently, version 1.10.0 to 1.6.1 and the latest development version are available with batch-tamarin. These Dockerfiles are available in the GitHub repository here. Ready to use Docker images are available on DockerHub: lmandrelli/tamarin-prover-and-batch

Docker images of the Tamarin prover only are also available on DockerHub: lmandrelli/tamarin-prover providing Tamarin Prover from version 1.4.0 up to 1.10.0, both for amd64 and arm64 architectures. The Dockerfiles used to build these images are available in the GitHub repository here.

Installation

Prerequisites:

Python 3.10 or higher
Tamarin Prover installed, preferably accessible in your system PATH

You can install batch-tamarin using pip:

pip install batch-tamarin

It will then automatically be available as a command line tool batch-tamarin.

Tamarin book published

Wed, 06 Aug 2025 08:00:00 +0000

Our Tamarin book is now available in hardcover from Springer Nature and electronically on our website.

Atom/RSS feed for new posts

Sun, 15 Jun 2025 08:00:00 +0000

Using a news feed reader? We now have an Atom/RSS feed available for new posts: Atom/RSS Feed.

Theory and exercise files for the book now available

Thu, 12 Jun 2025 19:00:00 +0000

You can now download the theory files and exercise files for the book on the book webpage. These files allow you to run the all example sources files from the book, and compare your exercise solutions to ours.

Website redesign

Mon, 09 Jun 2025 08:00:00 +0000

We are in the proccess of redesigning the Tamarin Prover website. Expect more updates in the next weeks as we work on the website.

Tamarin lecture slides

Fri, 16 May 2025 08:00:00 +0000

Aleksi Peltonen created a great lecture slide deck for a graduate-level course on Tamarin.

The slide deck is designed to be used together with the Tamarin book.

Use the slide deck to learn more about Tamarin, or use it to develop your own Tamarin course: the CC-BY license means you can reuse and adapt the material for your own purposes, as long as you attribute correctly.

The slide deck comes with the final PDF output but also includes the Latex source files used to produce them, so they can be easily adapted.

Tamarin book draft now available

Sat, 01 Mar 2025 08:00:00 +0000

We have been working on a book to support Tamarin users, which will be published soon by Springer. Learn more about our book here.