Some of the best-built software never reaches a single user. A founder runs out of runway a month before launch, a co-founder leaves, a bigger competitor ships first, or a day job takes over. The product is finished, or nearly, and it sits in a repository. For AI training and R&D, a launch is not what makes code useful: the engineering is. Odys AI Labs, the research and development arm of Odys, buys the source code of software that was never launched, for cash.
What your code is likely still worth
Unlaunched code is judged on the same factors as any other: size, originality, languages, quality, history, docs and tests. Users and revenue are not on that list. Our guide on how codebase value is measured explains each factor and how it is checked without anyone seeing your code.
Unlaunched products often have strengths people overlook. The history is usually clean and focused: months of commits building one product, without the noise of production firefighting. Design files tend to be complete, because the screens were designed before the build. And because no real users ever signed up, there is usually far less personal data to clean out than in a product that ran for years.
Marketplaces see it differently. Even Acquire.com’s pre-revenue exception requires the product to still be live and working, with a public website where users can sign up and log in, and its help pages say “coming soon” startups are rejected as listings. That is reasonable for platforms built to sell running businesses to buyers who want to keep them running. It also means an unlaunched codebase usually fits a direct buyer of code better. We do not publish price figures; the offer comes after review, and how we value a codebase is explained on the call.
Suppose two founders spent eighteen months building a property management platform: a TypeScript back end, a tenant mobile app, an owner dashboard, payment and document-signing integrations, and a few hundred tests. Their seed round fell through two months before the planned launch, and both took jobs. Nobody ever signed up. As an illustration, that is a complete product with a clear history and almost no personal data, which is a good starting point for a sale.
What a buyer will check
- Completeness. Can it be deployed and used end to end, or is it a set of screens and a partial API?
- Original share. How much your team wrote, versus starter kits, boilerplates, UI templates and generated code.
- History. Real commits over time, not a single upload.
- Rights. Who wrote it, and whether a company or individuals own it.
- Never earned. Confirmation that it was never sold or licensed to anyone.
Three risks specific to this situation
1. It is closer to a prototype than a product. This is the main question for unlaunched code. A repository with a landing page, a login screen and two half-built features is a prototype, and we pass on snippets. A product that a person could install and use, even with rough edges, is different. Be honest with yourself about which you have, and describe it plainly on the form. If the core is there but the edges are missing, say so: that is common and fine.
2. Too much of it came from templates or generators. Unlaunched products are often started from a SaaS boilerplate, an admin template or a framework scaffold, and in recent years with AI coding assistants generating large parts. None of that is wrong, but repos that are mostly copied libraries, templates or generated code are on our pass-on list. What matters is how much your team wrote on top. Our guide on counting lines of code shows how to separate original code from the rest with cloc or tokei.
3. Nobody ever sorted out ownership. Products that never launched often never got the paperwork either: no company formed, no founder agreements, no assignments from the friend who built the API. Whoever wrote code owns it first unless they assigned it in writing, so every meaningful contributor matters. And if anyone built it while employed elsewhere, their employment contract may claim it. Our guide on who owns the code covers the usual fixes. This is not legal advice; a lawyer can confirm your case.
What to do this month
- Try a fresh deployment from the repository and note what works and what does not.
- List which parts came from templates, starter kits or generators, and which your team wrote.
- Name everyone who contributed code and what, if anything, they signed.
- Remove test accounts with real details, scraped sample data and any keys in .env or config files.
- Gather the design files, specs and any planning docs or tickets, including exports from Jira, Linear or GitHub Issues if you used them.
- Write a one-page overview: what it does, the stack, how far it got and why it stopped.
Your options
| Option | For a product that never launched |
|---|---|
| Sell | Cash now; you cannot later launch this code as your own |
| Launch it | Worth trying if the reasons it stalled have changed |
| Open source | Can help your profile, but then it cannot be sold exclusively |
| Keep it private | Costs nothing; the stack ages and the option slowly fades |
| Delete | Removes every option, including a sale |
The choice between selling and launching is the real one here. Our comparison of selling your code versus relaunching it yourself sets out a plain decision rule.
If you sell, we never use your brand or name and never launch the product as yours. Odys AI Labs uses the code for AI training and R&D after cleaning, and we may also work on it with research partners. No code changes hands before a signed written agreement, and we never ask you to install or run anything.
What to do next
- Test whether the product deploys and works end to end.
- Get written agreement from everyone who wrote a meaningful part.
- Send us a few details for a free code valuation, or start with the code value check.
Frequently asked questions
Can I sell code for a product that never had a single user?
Yes. Never launched is on our list of situations that qualify. What we look for is a complete product, written by your own team, that makes no money today and that you own the rights to. Users and revenue are not required. Prototypes, snippets, templates and repos that are mostly copied libraries or generated code do not qualify.
How complete does it need to be?
Complete enough to be a real product, not a demo. A good test is whether someone could deploy it from the repository and use the main features end to end, even if it is rough. Missing polish, a half-finished admin panel or no marketing site is normal for an unlaunched product.
We never formed a company. Who owns the code?
Usually the people who wrote it, each owning what they wrote, unless they signed something that says otherwise. To sell, everyone with a meaningful share should agree in writing, or assign their rights to one person or entity first. This is not legal advice; a lawyer can confirm the right paperwork for your case.
Why not just launch it now instead of selling?
You can, and if you have a real plan to reach customers, it may be the better choice. Selling makes sense when the reasons it never launched still apply, such as no time, no budget or a market that moved. Once sold, the code cannot be launched by you as your own product.
Find out what your old code is worth right now.
Tell us about the product in about a minute. No code needed. We review the details and come back with a cash offer or a plain no.
Get my free code valuation →