Google Summer of Code 2026

[SaFiNguyen]

Hi Bob , I’m really interested in Gsoc 2026 and want to be a part of it.

Since GSoC is often considered beginner-friendly, I was wondering what criteria Synfig team usually use when evaluating applicants. In particular, what makes a strong proposal in your opinion?

Are there specific aspects you (or the mentors) pay the most attention to, such as technical depth, prior contributions, implementation plans, or something else? (^- ^ )

I’d really appreciate any advice on what applicants should focus on to write a high-quality proposal.

[BobSynfig]

For everone wanting to apply, I would like to share one of the main concerns among admins and mentors in GSoC email exchanges: AI and everything around this trend.

This has nothing secret but I will publish only the Subjects:

• [GSoC Mentors] Fun of being a mentor is significantly reducing with AI
• [GSoC Mentors] AI for admins and mentors
• [GSoC Mentors] Notes from today’s Discussion on Managing AI Spam and the Influx of Applicants
• [GSoC Mentors] Thoughts on requiring video proposal walkthroughs?
• [GSoC Mentors] How are people handling obvious copy/paste or LLM generated proposals because I saw at least 2, possibly 3.
• [GSoC Mentors] We need to have deep, fundamental discussions about GSoC goals, incentives, alignment and roles

There are a few others but as you can see, a lot are about AI “plague” and and the loss of purpose of GSoC (which is to wake up some interest for contributing to open-source for people who never did it before, including professional devs having only worked on closed-source)

There are also guidances (Vital Info and for all GSoC 2026 Mentors and Org Admins) which could look like funny but are indeed true:

Managing Contributor Inquiries:

You can expect a very high volume of inquiries as GSoC ramps up. Use these guidelines to manage the influx effectively while maintaining your sanity and sleep schedule:
(…)

Mentoring in GSoC requires significant effort, and burnout is a major concern.

Helpful Links

How to avoid, identify, and quickly sort through AI generated contributor proposals

So don’t forget to read and understand fully the rules, duties, implied things and expectations of enrollement in this program.
The temptation of AI is a threat for the spirit of GSoC, if used or abused, it could lead to the end of such a program.

And as I say every year, you can still contribute even though you are not taking part to GSoC

[rodolforg]

Please understand:

• GSoC is about bringing people to Free/Open Source communities - to make code contributions and - we hope - keep contributing to open source ecology after the GSoC period;
• If you insists/wants to use an AI tool for coding for you, maybe you should go somewhere else:
  • it’s really boring/annoying to review AI-generated code for many reasons
    • specially if the tool user doesn’t understand the slopped code - we cannot help him/her to understand what s/he didn’t write
    • verbose
    • codes that do not do as the tool commented right above
    • often does not compile or doesn’t work
  • if you don’t invest/spend time studying and writing code, why should I do it for you?
  • how can we know you are not infringing any copyright license if the AI-tool is essentially made from (not always authorized) copyrighted code? Our code is GPL-licensed and we shall obey this license.
  • …

Please don’t believe it is a whole new world era. This stuff exists for 10 years at least, there was even a TED talk about it then…

[ahmed1]

Hello Bob,@mohamed.Adhamc, @rodolforg and everyone else! My name is Ahmed and I am a third-year Computer Science student with a passion for writing maintainable software (and I mean so genuinely ). Recently, I was informed of the great opportunity which is Google SoC by an older mentor who is the biggest open source enthusiast you will ever meet. I then spent a great deal of time reading about participating organizations that looked interesting, but I am happy to share I have now narrowed it down to Synfig. I hope to be of value to this widely cherished community. As someone who has dedicated a great portion of his time to contributing to a closed source Biomedical research project I have consistently held an intriguement for the light side: open source software. I see contributing as a great way to bring about net positivity to our world and also engage in intellectual stimulation. Expect to see more of me around these parts as I intend to be here for the long-run: I know very well that “Art is never finished, only abandoned.” (Leonardo Da Vinci?)

[ahmed1]

@BobSynfig @rodolforg As someone who is still technically mostly an outside to the world of open source, I, at least, have always viewed it to be a pure and thoughtful endeavor. I share the sentiment that AI-generated code is out of place in the code base, and shouldn’t actually make it there in the first place since it is simply not worthy of the time of a human reviewer. I can not deny transformer based models such as LLMs may have some utility since an increasing number of people appear to be increasingly accepting its outputs as a source of valuable truth. However, having played around with agentic workflows as much as I have (and consequently observing their stark imperfections) and understanding the nature of the underlying statistical processes which power these models, I will forever hesitate to admit these models possess anything remarkable other than hype.