How to Get Your GSoC Project Accepted: Lessons from Rust’s 2026 Selection

By ⚡ min read

Overview

Every year, Google Summer of Code (GSoC) offers a unique opportunity to dive into open source. The Rust Project’s 2026 cycle was no exception—with 96 proposals submitted and only 13 projects accepted, the competition was fierce. This guide unpacks exactly how the Rust mentoring team evaluated proposals, and what you can do to stand out. Whether you’re targeting Rust or another organization, the principles here—early engagement, genuine contributions, and a well-structured proposal—are universal. By the end, you’ll have a clear roadmap to craft a proposal that mentors can’t ignore.

How to Get Your GSoC Project Accepted: Lessons from Rust’s 2026 Selection — Source: blog.rust-lang.org

Prerequisites

Before you start writing your proposal, ensure you have:

Basic familiarity with the project’s domain – for Rust, that means comfort with at least one of: compiler internals, tooling, library design, or specific domains like GPU offloading or WebAssembly.
Access to the project’s communication channels – Rust uses Zulip for GSoC discussions. All project ideas are posted there, and mentors actively engage with applicants.
A GitHub account – to make contributions and submit patches. Many accepted contributors had already submitted merges before the official start.
Time and motivation – GSoC is a full-time commitment (12 weeks). Proposals that show deep interest and prior work are far more likely to succeed.

Step-by-Step Guide to a Winning Proposal

Step 1: Study the Project Ideas List

Rust’s GSoC organizers publish a curated list of project ideas months in advance. For 2026, these ranged from “A Frontend for Safe GPU Offloading in Rust” to “Implementing impl and mut restrictions.” Read each carefully. Note which align with your skills and interests. The best proposals come from genuine excitement—not from trying to force a fit. As the Rust team observed, “We had to pick only one proposal per project topic,” so choose wisely.

Step 2: Engage Early on Zulip (Before March)

Don’t wait until the proposal deadline. Mentors from the Rust Project start discussing ideas with applicants as soon as they’re published. The 2026 cycle saw “many interesting discussions” on Zulip, and some contributors even made non-trivial contributions to repositories before GSoC officially started. How to do it:

Introduce yourself in the relevant stream (e.g., #gsoc or #project-name).
Ask clarifying questions about the project scope, expected outcomes, and mentor expectations.
Share your initial thoughts or rough timelines. Mentors appreciate proactive candidates.

Step 3: Make Non-Trivial Contributions

Proposals backed by real code stand out. The Rust team explicitly noted that “some applicants made non-trivial contributions to various Rust Project repositories” before the deadline. These contributions demonstrate both skill and commitment. Aim for at least one meaningful pull request—bug fix, feature addition, or documentation improvement—related to your chosen project area. Even if it’s not merged, the mentor will see your effort.

Step 4: Craft a Detailed, Personalized Proposal

With 96 proposals, generic ones get lost. Here’s what to include:

Problem statement: Why does this project matter to the Rust community? Reference the original idea but expand with your own research.
Technical approach: Outline the architecture, languages, tools you’ll use. Be specific. For example, for the “Debugger for Miri” project, you’d discuss how to extend Miri’s internals.
Timeline: Break into phases (e.g., weeks 1–4: research + prototyping; weeks 5–8: core implementation; weeks 9–12: testing + documentation).
Prior interactions: Mention your Zulip discussions and any contributions. This shows you’ve already built rapport.
Importance to community: Explain how the project aligns with Rust’s wider goals (e.g., safety, performance, ecosystem growth). The mentors evaluate “the importance of the proposed project for the Rust Project and its wider community.”

Step 5: Be Mindful of Mentor Bandwidth

In 2026, “some projects had to be canceled due to several mentors losing funding for Rust work.” Even before that, mentors couldn’t take on more than one or two projects each. If you propose a project that requires heavy mentorship, it might be deprioritized. Show that you can work independently while still seeking guidance. Avoid overloading a single mentor: check who the listed mentor is and whether they are already assigned to other projects.

Step 6: Submit Early—and Avoid AI Pitfalls

Many organizations, including Rust, struggled with “AI-generated proposals and low-quality contributions generated using AI agents.” While tools can assist, a proposal that reads like it was written by a bot is a red flag. Ensure your proposal reflects your voice and understanding. Submit a few days before the deadline so you can address any mentor feedback if they request revisions.

Common Mistakes to Avoid

Mistake 1: Not Interacting with the Community Beforehand

Proposals from complete strangers often feel cold. The Rust team gave strong weight to “prior interactions with the given applicant.” Without any Zulip history or GitHub contributions, your chances drop significantly.

Mistake 2: Ignoring the Project’s Wider Impact

Mentors evaluate “the importance of the proposed project for the Rust Project and its wider community.” If your proposal only benefits a niche corner, it might be passed over. Tie your work to Rust’s strategic priorities (e.g., GPU offloading, WebAssembly, safety improvements).

Mistake 3: Overpromising or Being Vague

A proposal that says “I will implement feature X” without explaining how is weak. Conversely, promising something huge in 12 weeks looks unrealistic. The accepted proposals for 2026 had clear, achievable scopes. For instance, “Adding WebAssembly Linking Support to Wild” is a well-defined deliverable.

Mistake 4: Submitting an AI-Generated Proposal

Rust reported that “AI-generated proposals and low-quality contributions generated using AI agents” were a problem. While not automatically disqualifying, they often lack depth. Mentors can spot generic language. Write your own.

Mistake 5: Ignoring Mentor Availability

Even a brilliant proposal can be rejected if the mentor cannot take it. The Rust team had to “cancel some projects due to several mentors losing their funding.” Check the mentor list on Zulip and maybe ask if they have capacity. If they are overloaded, consider a different project.

Summary

Getting a GSoC project accepted by the Rust Project—or any major open-source organization—requires early, genuine engagement and a proposal that balances technical depth with community impact. In 2026, only 13 out of 96 proposals (13.5%) were selected. Those that succeeded started conversations on Zulip, made non-trivial contributions, aligned with mentor bandwidth, and avoided AI-generated shortcuts. Follow these steps: study the ideas, engage early, contribute code, craft a personalized proposal, and respect mentor limits. You’ll dramatically increase your odds. Good luck!