MTurk Alternative: Where to Move After the Shutdown

Par Priya Raman · Mis à jour le 2026-11-08

This site earns a commission if you sign up through its links (Hitworkers affiliate program). Code CHICO5 gives you +5% credits and reduces the commission from 12% to 7% on your account.

Amazon Mechanical Turk shut down permanently on September 30, 2026, so requesters now need to split their old MTurk workload across a few different tools depending on the task. Simple paid micro-tasks go to marketplaces like Microworkers, SproutGigs or Hitworkers, while thumbnail and title testing has a dedicated free option inside YouTube Studio.

En bref

I ran labelling and QA tasks through MTurk on and off since 2017, mostly small batches for data cleanup and survey screening. When Amazon pulled the plug on requester.mturk.com on September 30, 2026, a lot of people in my Slack channels were asking the same question: where do I send this work now. The honest answer is there is no single drop-in replacement. It depends on what the task actually is.

Why did MTurk shut down, and does it matter for my migration?

The reason doesn't change what you need to do next, so don't spend too much time on it. What matters is that the requester-side tools, templates and qualifications are gone, so any workflow built around MTurk's API needs a new home.

If you were running simple paid micro-tasks (click, sign up, rate, review), that workload maps cleanly onto the micro-task marketplaces below. If you were running structured data labelling at scale, that's a separate category of tool entirely and outside what this page covers, since none of my verified sources cover enterprise labelling platforms.

What did MTurk actually cost, for comparison

MTurk charged requesters a 20% fee on top of worker payouts, and that jumped to 40% once a batch went over 10 assignments. That fee structure is useful as a baseline when you're reading pricing pages on other platforms, because "fees" means different things on different sites.

Fee comparison table

PlatformFee basisRateMinimum to start
Mechanical Turk (closed)On all assignments20%, 40% above 10 assignmentsn/a, shut down
MicroworkersOnly on tasks rated "Satisfied"10% (7.5% under $0.10/task)$10 deposit
SproutGigsOnly on satisfactory tasksApproval fee, see pricing pageCountry targeting available
HitworkersIncluded in displayed task price10%, shown upfrontCampaign example shown: 30 workers x $0.30

The pattern across the newer platforms is the same: you're billed on approved work, not on everything you post. That's a meaningful difference from MTurk's flat 20% on every assignment regardless of quality.

Where should simple paid micro-tasks go now?

For things like account sign-ups, app installs, surveys, or quick reviews, the three platforms worth testing are Microworkers, SproutGigs and Hitworkers. Each has its own proof and approval flow, so the right pick depends on how fast you need results and how much control you want over campaign setup.

Microworkers runs on "Standard" and "TTV" campaign types and only bills the approval fee on tasks marked Satisfied, with a $10 minimum deposit to get started, per their own FAQ page. If you're coming from MTurk's qualification system, their long track record (they're one of the oldest names in this space) makes the learning curve shorter. For a deeper side by side, I've covered this in Sites Like Microworkers: 6 Platforms Compared and broken the pricing down further in Microworkers Fees vs Competitors: True Cost per Task.

SproutGigs leans on country-level targeting for micro-jobs and, like Microworkers, only pays out on tasks that pass review, with its own approval fee listed on its pricing page. If targeting a specific country's workers was a big part of your MTurk setup, that's worth a closer look, and I go into it in SproutGigs Alternative: Better Proof, Faster Approval?.

Hitworkers is newer (the domain went live October 21, 2026, and the company behind it, MEDIASCALE LLP, was registered in England in September 2026) but it's built specifically around proof-based campaigns with a 5-step setup: title, instructions, price, audience, launch. The pricing example shown on their dashboard is 30 workers at $0.30 each for $9.90 total, with the 10% platform fee already included in that number. They currently show a temporary offer letting you launch a campaign with as few as 10 workers instead of the usual 30, for 7 days. Templates cover things like watching and subscribing on YouTube, posting on LinkedIn, searching on Google, creating Gmail accounts, or publishing on YouTube, plus public task types like forum link placements (from $0.50 per link, with proof required as URL, login and screenshot) and thumbnail testing votes (from $0.10 per vote, 2 to 4 thumbnails, real viewers asked to justify their pick). Workers come from 190+ countries, there's no subscription, and rejected work gets refunded. On the worker side, withdrawals carry a 20% fee, which is worth knowing if you ever get asked about it by task-takers.

Launch a campaign →

One detail I want to be upfront about: Hitworkers runs an affiliate program, and I earn 12% of what an advertiser spends on validated tasks if they sign up through my link, paid for life on that account. They also have a promo code, CHICO5, which gives you (the advertiser) an extra 5% in credits on your payments. The catch is that using that code drops my commission on your account from 12% down to 7%. I'm telling you this because you should know exactly who benefits from which choice: the code is better for you, the plain link is better for me. Use whichever makes sense for your situation.

What if my old MTurk tasks were thumbnail or title testing?

If you were paying MTurk workers to vote on thumbnails or titles, YouTube now has a built-in free tool for that called Test & Compare, inside YouTube Studio's advanced features. It tests up to 3 titles and/or thumbnails directly on real viewers and picks a winner based on watch time share, over a run that can last a few days to two weeks.

The tradeoff is that Test & Compare needs a published video with real impressions, and it won't tell you why one thumbnail won, just which one performed better on actual watch time. A paid micro-task platform like the thumbnail voting option on Hitworkers gives you qualitative feedback (workers explain their pick) before you publish, which is a different kind of signal. Many creators now use both: quick opinion-based testing before launch, then Test & Compare after. I compare the proof quality and turnaround of a few of these task marketplaces in Hitworkers vs Microworkers: Head-to-Head.

What about QA, link building or SEO-adjacent tasks I used to run on MTurk?

These were always a gray area on MTurk, and they still require care wherever you run them, since manipulating search signals or click-through rates can conflict with a search engine's spam policies. If you're moving forum link building or review-style tasks to a new platform, check the fee structure and proof requirements first, since not every marketplace handles this type of task the same way.

If you're running a broader QA or crowd-review program rather than one-off tasks, it's worth comparing a few platforms side by side before committing, which is exactly the kind of decision I walk through in Crowdsourcing Platform for Businesses: Buyer's Guide. And if your volume is small and you don't want to manage multiple accounts, Clickworker Alternative for Small Campaigns covers a lighter-weight option worth testing first.

So which one should I actually pick?

There's no universal answer, but a simple rule works for most people migrating off MTurk: match the platform to the task type rather than trying to force everything into one tool. Micro-tasks with clear pass/fail criteria fit Microworkers or SproutGigs well because of their straightforward approval-based billing. Campaigns needing specific proof formats, faster setup, or a temporary low worker minimum fit Hitworkers. Thumbnail and title decisions belong in YouTube's native tool once you have a video ready to publish.

Start with a small test batch on whichever platform matches your task type, check how review and proof requirements compare to what you remember from MTurk, and scale from there once you've seen the approval rate for yourself. If you're still deciding between the main micro-task options generally, my overview at Best Microworkers Alternative for Advertisers (2026) is the right starting point.

FAQ

Is Amazon Mechanical Turk really shut down for good?

Yes, MTurk closed permanently on September 30, 2026, and the requester portal at requester.mturk.com is no longer operating. There's no announced relaunch, so any workflow built on it needs to move to another platform.

What's the closest one-to-one replacement for MTurk?

There isn't one single replacement since MTurk covered a wide range of task types. For simple paid micro-tasks, Microworkers, SproutGigs and Hitworkers are the closest matches, while thumbnail and title testing fits better inside YouTube's native Test & Compare tool.

How do the fees compare to what MTurk used to charge?

MTurk charged 20% on all assignments and 40% once a batch exceeded 10 assignments. Microworkers charges 10% (or 7.5% for tasks under $0.10) only on approved work, and Hitworkers includes its 10% fee directly in the displayed task price, so you're not billed on rejected work with either.

Can I still access my old MTurk task history or templates?

Based on what's publicly available, the requester portal is closed, so there's no indication templates or history remain accessible. Anything you relied on there will need to be rebuilt from scratch on a new platform.

What is the Hitworkers promo code CHICO5 and should I use it?

CHICO5 gives you, the advertiser, an extra 5% in payment credits on Hitworkers. It also lowers the commission I earn on your account from 12% to 7%, so I'm disclosing that tradeoff openly rather than hiding which option benefits me more.

Is YouTube's Test & Compare free to use?

Yes, it's a built-in feature of YouTube Studio's advanced features, not a paid add-on, though it does require identity verification steps on your channel and a published video to generate real watch time data.

Do any of these platforms handle large-scale data labelling like MTurk used to?

None of the sources I rely on here cover enterprise-scale data labelling specifically, so this page focuses on micro-tasks, surveys, QA and thumbnail testing rather than structured labelling pipelines.

Sources