Welcome, and thanks for your interest in contributing to PyMC through Google Summer of Code 2026! We’re excited by the level of enthusiasm we’re seeing. To keep things manageable for our small team of volunteer mentors, we’ve put together this guide to answer the questions we’re seeing repeated across many introduction posts. Please read this carefully before opening a new topic.
Available Projects for 2026
Full descriptions of the candidate projects, including expected outcomes, required skills, and mentor contact information, are on the GSoC 2026 projects wiki page.
Frequently Asked Questions
Q: I’m new to PyMC. What should I do first?
Start using the tools. This is the single most important thing you can do right now. Before you can help build PyMC or PyTensor, you need genuine hands-on experience using them to solve real problems. Pick a scientific or analytical question you’re genuinely curious about and build a Bayesian model for it in PyMC. Explore the example notebooks, work through Statistical Rethinking, or watch one of many recorded tutorials. If your interests are related to backend computation rather than Bayesian modeling, the same advice applies — but starting with PyTensor.
It is hard to be an effective toolmaker for these libraries if you aren’t an active user of them yourself. A notebook demonstrating that you’re genuinely exploring PyMC/PyTensor to address a problem you care about is the most compelling thing you can show us.
Q: Should I start by finding GitHub issues to close?
Not necessarily — and this is a common misconception. Working on issues is a useful way to learn the library, but it shouldn’t be your primary goal at this stage. A notebook showing that you’re actively using and exploring PyMC on a problem you care about is more valuable than a slew of trivial changes to the codebase.
Q: How do I engage with a potential mentor?
Open a topic in the Development category with a clear title indicating the project (e.g., “GSoC 2026: Survival Models”). Show your work: share what you’ve explored in the codebase, any experiments or notebooks you’ve produced, and your initial thinking on the project. Mentors are more likely to engage substantively when you’ve clearly done your homework, and not relied on LLM output to feign understanding.
Please don’t cross-post across multiple threads or repeatedly tag core developers — our mentors are volunteers with limited time, and doing so creates a negative first impression.
Q: How many GSoC slots will PyMC receive?
We don’t know in advance. Google allocates slots based on the quality and quantity of submitted proposals, and the number varies year to year. The selection process is competitive. Our best guess is 1-4 slots, based on past years, but this is not guaranteed.
Q: What makes a strong proposal?
A strong proposal demonstrates:
- That you understand the problem you’re solving and why it matters to the project
- That you’ve explored the relevant parts of the codebase and engaged with the mentors
- A realistic, week-by-week work plan with defined deliverables and milestones
- Awareness of technical risks and open questions, with a credible approach to addressing them
- Evidence that you are actively using PyMC/PyTensor (notebooks, prior PRs, Discourse engagement)
Q: I want to send a draft proposal for feedback. Who should I contact?
Once you have a working draft, send it to fonnesbeck+gsoc2026@gmail.com for review. Include your Discourse username and the project title in the subject line. Early drafts shared well before the March 31 deadline are more likely to receive feedback than last-minute submissions.
Q: I’m interested in long-term involvement beyond GSoC. Is that possible?
Absolutely, and we encourage it. Some of our most valued contributors started through GSoC and became long-term members of the community. The best way to build toward that is to show up consistently, engage thoughtfully, and produce well-documented, maintainable work.
Key Dates
| Date | Milestone |
|---|---|
| Mar 16 | Contributor application period opens |
| Mar 31 | Contributor application deadline |
| Apr 30 | Accepted contributors announced |
| May 1–24 | Community Bonding Period |
| May 25 | Coding officially begins |
| Jul 6–10 | Midterm evaluations |
| Aug 17–24 | Final work submissions |
We are currently in the discussion window. This is the time to engage with the community, start using the tools, and begin drafting your proposal.
We look forward to seeing what you build. Good luck!
— The PyMC Development Team