AI Product Evaluator

What you'll do

You test AI models and agents against scenarios that are already written for you. Each run is a sequence of operational tasks meant to probe accuracy, reasoning and responsiveness; you carry them out, note what the model did, and log anything odd in a structured report. Then you reset the usage limits and do it again, as many times as the protocol asks.

The ad is candid that this is operational work rather than creative evaluation. Deadlines are tight and often tied to when usage limits reset, so timing matters as much as judgment. Your findings and the resulting evaluation datasets go onto micro1's AI training platforms against project milestones.

Who fits

• Experience with AI products, models or agent testing
• The discipline to follow complex instructions precisely and identically across many runs
• Strong attention to detail and the habit of documenting each step
• Someone who already uses AI tools in daily life
• Comfort with repetitive, high-volume tasks on strict timeframes
• Digital product, software QA or AI research project experience is a plus

No degree or minimum years are stated.

What it pays

micro1 publishes $14–36/hour. The top is more than double the bottom and the ad does not explain what decides placement. With three very different labour markets eligible, location may play a part, but that is our reading rather than micro1's statement. No pay structure, weekly hours or project length are given.

Worth knowing

Good:

• 100 openings, flagged new and high demand
• A low bar to entry: no degree, no stated years, and everyday AI use counts
• Clear scripts, so there is little ambiguity about what a task wants
• A sensible first step into evaluation work for software QA testers

Less good:

• Only residents of India, Japan and the United Kingdom are eligible
• The $14 floor is low for UK-based contributors
• Repetitive by design; the same scenarios run again after each reset
• Strict timeframes tied to limit resets can make the schedule less flexible than typical task work
• No hours, duration or start date published

About this listing

Posted by micro1 on its referral jobs board and read on 18 October 2026. The pay band, the 100 openings, the three-country location line and the qualifications are micro1's own text. The customer and the products under test are not named, and payment schedule and visa terms are not published on the listing.

Compared with micro1's AI Agent Evaluator ($25–45/hour), this listing pays less at the bottom and leans more on executing fixed test protocols than on open-ended judgment. See micro1.