
Open Source GitHub Maintainer
Posted 14 Aug 2026
Role Title: Open source GitHub maintainer Role Type: Contractor (15 hrs a week) Location: Remote, Open Globally Schedule: Flexible, any hours/days of the week. As long as there is…
Every open freelance coding job from seven AI training marketplaces, highest hourly rate first. No proposals to write and no clients to chase. Refreshed daily.
178 open freelance coding listings live right now, paying up to $300 an hour.
Filtering by rate hides per-task listings, which have no hourly equivalent.

Posted 14 Aug 2026
Role Title: Open source GitHub maintainer Role Type: Contractor (15 hrs a week) Location: Remote, Open Globally Schedule: Flexible, any hours/days of the week. As long as there is…

Posted 16 Sep 2026
Pay: $100–$150/hour Location: Global, fully remote Job Type: Contractor (~15 hours per week) Schedule: Flexible—you choose the hours and days you work, including weekends if needed…

Software Engineer Overview The newest models can vibe-code a whole feature in minutes. Someone has to check whether that code actually works, and that someone is you. You'll run ag…

Posted 3 Nov 2026
Job Title: Solar2D Game Developer Job Type: Contractor Location: Remote micro1 is looking for experienced Solar2D developers to support an AI training project. In this role, you wi…

Posted 3 Nov 2026
Job Title: Software Engineer (Open 3D Engine) Job Type: Contractor Location: Remote Job Summary: micro1 is looking for experienced Open 3D Engine (O3DE) specialists to support an A…

Posted 3 Nov 2026
Role Title: 2D Game Developer (Defold) Role Type: Contractor Location: Remote micro1 is looking for experienced Defold developers to support an AI training project. In this role, y…

Posted 3 Nov 2026
Job Title: Python Game Developer (Panda3D) Job Type: Contract Location: Remote micro1 is looking for experienced Panda3D developers to support an AI training project. In this role,…

Posted 16 Oct 2026
Puzzle Solver (Coding) Pay: $60–$120/hour Location: Global, fully remote Job Type: Contractor (~15 hours per week) Schedule: Flexible—you choose the hours and days you work, includ…
Posted 7 Sep 2026
Software Engineer – Frontend (AI Training Contract) About the Role Frontend developers: your expertise is in demand — not for building products, but for making AI smarter. We're lo…
Posted 4 Sep 2026
Software Testing Analyst (AI Training) About the Role What if your coding instincts could directly influence how AI writes software for millions of people? We're looking for Softwa…
Posted 4 Sep 2026
Code Review Specialist (AI Training) About the Role What if your coding knowledge could directly influence how the next generation of AI writes software? We're looking for Code Rev…
Posted 3 Sep 2026
Software Engineer Task Author (AI Training) About the Role Alignerr is building a dataset of expert tasks that train and evaluate advanced AI agents on real enterprise engineering…
Posted 3 Sep 2026
SWE - Backend (Contract) Labelbox • Remote (United States preferred) Shape the data that powers frontier AI --- Quick facts Engagement | Hourly, at‑will contractor Schedule | Fully…
Posted 3 Sep 2026
SWE - SQL (Contract) Labelbox • Remote (United States preferred) Shape the data that powers frontier AI --- Quick facts Engagement | Hourly, at‑will contractor Schedule | Fully rem…
Posted 2 Sep 2026
Coding Expert - JavaScript (Contract) Labelbox • Remote (United States preferred) Shape the data that powers frontier AI --- Quick facts Engagement | Hourly, at‑will contractor Sch…
Posted 2 Sep 2026
Coding Expert - Java (Contract) Labelbox • Remote (United States preferred) Shape the data that powers frontier AI --- Quick facts Engagement | Hourly, at‑will contractor Schedule…
Posted 26 Jun 2026
About the Role Alignerr is building a dataset of expert tasks that train and evaluate advanced AI agents on real enterprise work. As a Task Author for the Software Engineer role, y…
Posted 28 May 2026
Senior Software Engineer – AI Evaluation About the Role What if your engineering skills could directly determine whether the world's most advanced AI systems are actually working?…
Posted 3 Mar 2026
Coding Expert – C++ (AI Training) About the Role We're looking for experienced C++ developers to help evaluate and improve AI systems that write and reason about code. Your experti…
Posted 3 Mar 2026
Golang Coding Expert (AI Training) About the Role We're looking for experienced Go (Golang) developers to help train and evaluate the next generation of AI coding models. Your expe…
Posted 3 Mar 2026
Coding Expert – Rust (AI Training) About the Role We're looking for experienced Rust developers to help evaluate and improve AI systems that generate and reason about code. Your ex…
Posted 3 Mar 2026
DevOps / IaC Engineer (AI Training Contract) About the Role What if your hard-won DevOps expertise could directly shape how AI understands and reasons about cloud infrastructure? W…
As an Android Engineer (Kotlin / Java), you will work remotely on an hourly paid basis to review AI-generated Android code, mobile architectures, and technical explanations, as wel…
As an iOS Engineer (Swift / Objective-C), you will work remotely on an hourly paid basis to review AI-generated iOS code, mobile architectures, and technical explanations, as well…
These freelance coding jobs come from AI labs, not clients. The labs pay developers by the hour to write, review and test the code their models learn from. Most listings publish an hourly rate before you apply. Nobody asks for a proposal and nobody haggles.
That is the difference from Upwork, where you buy Connects to bid and a study of 44,863 postings found 6 in 10 never hired anyone. Here you apply once, pass a screen, and take tasks as they open.
There is one catch: the screen. Most of these marketplaces interview you with an AI that follows up on every answer, then grade a short sample of the real work. Strong engineers fail it by coding in silence. It scores how you explain a decision, not only whether the code runs.
Published hourly rates for each section on this page, computed from the open listings. A listing counts in more than one row when its title names several.
| Stack or specialty | Open listings | Published hourly rate |
|---|---|---|
| Software engineer | 41 | $6 to $150 |
| GitHub contributor | 6 | $20 to $300 |
| Python | 11 | $50 to $120 |
| JavaScript and TypeScript | 18 | $20 to $120 |
| Rust | 12 | $20 to $120 |
| C++ and C# | 15 | $20 to $120 |
| Code review and QA | 13 | $40 to $120 |
Freelance software engineer jobs are the general seats: listings that name the role rather than a language. The work spans your whole stack, from fixing a model’s pull request to writing the test that proves it wrong.
Seniority sets the rate here more than anywhere else on the page. Junior, senior and specialist titles from the same marketplace can sit far apart, so check the level a listing asks for before you apply.
41 open software engineer listings right now, published at $6 to $150 an hour.


Posted 3 Nov 2026
Posted 7 Sep 2026
Posted 3 Sep 2026
Posted 3 Sep 2026
Posted 26 Jun 2026
Most open source maintainers work unpaid. These listings pay engineers by the hour to work in public repositories: fix real bugs, build features, and write the tasks and tests that check whether an AI agent can do the same.
The screen reads your public record, so expect to walk through a pull request you shipped and the review it got. One listing spells out the bar: contributions must show real engineering, not only documentation updates, dependency bumps or typo fixes.
6 open GitHub contributor listings right now, published at $20 to $300 an hour.

Posted 14 Aug 2026

Posted 16 Sep 2026

Posted 14 Oct 2026

Posted 14 Oct 2026

Posted 14 Sep 2026
Freelance Python jobs lead this market, because coding models are tested on Python more than any other language, and it is the skill these listings name most often. SWE-bench, the coding-agent benchmark labs quote most, draws every task from Python repositories.
The listings ask for senior backend and full-stack Python: typing, packaging, concurrency and the bugs a test suite misses. You grade a model’s solution, fix it, and write the tests that prove the fix.
11 open Python listings right now, published at $50 to $120 an hour.

Posted 3 Nov 2026
Posted 3 Sep 2026
Posted 3 Sep 2026
Posted 2 Sep 2026
Posted 2 Sep 2026
Posted 7 Sep 2026
Frontend and Node work, reviewed the way a senior engineer reviews a pull request. You catch the component that renders but breaks on resize, the effect that fires twice and the type that compiles but lies, then write the fix.
Expect React, TypeScript and Node.js in the requirements. Full-stack titles in other languages sit in their own sections.
18 open JavaScript and TypeScript listings right now, published at $20 to $120 an hour.
Posted 7 Sep 2026
Posted 2 Sep 2026
Posted 3 Sep 2026
Posted 2 Sep 2026
Rust listings sit at the systems end of the market: data infrastructure, distributed systems, ML tooling and high-performance computing. Most titles ask for senior, lead or principal experience.
The tasks match the language. You review a model’s Rust for ownership, concurrency and unsafe-code mistakes, then write the version that passes the borrow checker and the tests.
12 open Rust listings right now, published at $20 to $120 an hour.
Posted 3 Mar 2026
Posted 2 Sep 2026
Posted 2 Sep 2026
Posted 3 Mar 2026
Posted 3 Mar 2026
Posted 3 Sep 2026
C++ listings cluster around performance: HPC, data platforms and internal tooling. C# listings are .NET backend and full-stack work.
Both reward reading code before running it. Models still write C++ that compiles and then leaks, races or allocates in a hot loop. Your job is to spot it and explain why.
15 open C++ and C# listings right now, published at $20 to $120 an hour.
Posted 3 Mar 2026
Posted 7 Sep 2026
Posted 4 Sep 2026
Posted 3 Sep 2026
Posted 2 Sep 2026
Posted 2 Sep 2026
Code review is the skill AI labs buy most directly. These listings pay you to judge code you did not write: rank two model answers, audit an agent’s trace step by step, or test a feature until it breaks.
Auditor roles check other contributors’ tasks against a benchmark’s rules, SWE-bench among them. Pairwise and ranking roles compare two solutions and explain which one should merge. It is the closest match to a senior engineer’s day, without the meetings.
13 open code review and QA listings right now, published at $40 to $120 an hour.
Posted 4 Sep 2026
Posted 4 Sep 2026
Posted 28 May 2026

Posted 14 Sep 2026
Posted 22 Jun 2026
Posted 28 Aug 2026
A task is a prompt, a model’s answer and a rubric. You run the code, find what is wrong, explain each problem, and often write the version that should have shipped. The better-paid tasks ask you to build the test itself: a small repository, a planted bug and a suite that proves whether an agent fixed it.
You choose when to work. Most listings are hourly contracts with no meetings, so tasks fit around a day job or a gap between clients. Some pay for hours logged and some for accepted tasks. The listing says which, so read it before you start.
Seven marketplaces publish this work openly. Every listing above links to the one that posted it, and each has its own page here covering pay, screening and who gets hired.
Yes. You join the marketplace as an independent contractor, and most listings publish an hourly rate up front. There is no client to find, no scope to negotiate and no invoice to chase. The marketplace matches you to projects after you pass its screen.
The screen. An AI interview or coding assessment scores how you explain decisions, not only whether the code runs, and engineers who think in silence fail it. Plan around volume too: projects open and close in waves, so hours are not guaranteed.
On Upwork you bid for each project with paid Connects and pay a service fee of 0% to 15%. Toptal screens you, then matches you with clients you still have to win. Here most listings publish the rate up front and nobody bids. The tradeoff: you build training data for AI labs, not products for clients.
The rates table on this page answers that from today’s listings. Seniority moves the rate more than language does: on the same marketplace, a staff engineer auditing agent traces earns a multiple of what entry coding tasks pay.
No. The labs employ ML researchers. They pay you for engineering judgment: reading code critically, writing tests that catch real bugs and explaining a fix in plain English.
Usually, if your employment agreement allows paid outside work. You produce training data, not a product, so invention assignment clauses rarely reach it. Confidentiality runs both ways: never paste employer code into a task, and never reuse project material elsewhere.
AI training pays for judgment, not shipped product. If you want client projects, these are the established routes.
Every marketplace above screens with an interview or assessment in your own domain before assigning paid work. Run that interview with Skillora first and get feedback on how you explained your reasoning, not just on what you said.
Practise the screen free