How Much I Really Earned in
My First 3 Months Training AI

Most "I made $X training AI" posts online either feel suspiciously vague or suspiciously perfect. So here's a genuinely detailed breakdown of my own first three months — including the slow weeks, the platform that barely paid off, and the actual numbers, not rounded-up estimates.

For context: I worked part-time around an existing schedule, roughly 8–12 hours per week, split across two to three platforms at any given time. This isn't a "quit your job" story — it's what a realistic side income actually looks like.

Month 1 is the hardest month in AI training — not because the work is hard, but because the approval and matching pipeline takes time that feels like wasted time. It is not. Every platform you apply to in Month 1 is potentially paying you by Month 2.

Month 1: Getting Set Up

The first month was mostly setup, not earning. I applied to Mercor and SME Careers simultaneously in week one. Mercor's interview took about 20 minutes; SME Careers was just a short form.

Month 1 total: roughly €350, across about 20 hours of actual work, not counting the unpaid time spent on applications and onboarding.

Month 2: Adding a Third Platform

By month two, I had a feel for how inconsistent task availability could be, so I added micro1 as a third platform specifically to smooth out the slow weeks. I also started setting a higher self-declared rate on micro1 once I had some completed work to point to.

Month 2 total: roughly €890 — more than double month one, mostly from having three platforms covering each other's slow weeks rather than from any single breakthrough.

Month 3: Finding a Rhythm

By month three, the biggest improvement wasn't a new platform — it was simply understanding which task types paid best on each platform and prioritizing those.

Month 3 total: roughly €1,000 — almost triple the first month's earnings, on a similar weekly time commitment.

The Full 3-Month Summary

Total across 3 months: roughly €2,240 for about 93 hours of actual work. The effective hourly rate climbed steadily — not because any single platform suddenly paid more, but because spreading across multiple platforms reduced wasted idle time, and a track record on each platform gradually unlocked better-paying tasks.

What I'd Do Differently

• Apply to all three platforms in week one, not staggered. I lost income in month one simply by not having micro1 approved yet.
• Set up Wise before the first payment arrives. I used PayPal for the first few payments and lost a noticeable amount to conversion fees before switching.
• Don't underprice on micro1 out of caution. I started too conservative and left money on the table for the first few weeks.

Is This Realistic for You?

This was a part-time, side-income approach — roughly 8–12 hours per week. It's not a full-time replacement income at this pace, and it wasn't trying to be. If you're considering this as a serious side income or eventually a larger part of your earnings, the pattern that worked here — multiple platforms, consistent weekly effort, gradually building a track record — is the realistic path, not a single lucky break.