What the work involves

You review outputs from models that generate or assist with 3D content: text-to-3D meshes, motion sequences, retargeted animation, material and shader setups, and rendered stills. A typical task gives you a prompt, one or more model outputs, and a rubric. You decide which output is better and write the reasoning — not "the walk cycle looks off" but that the contact pose has no weight shift, the hips don't counter-rotate, and the foot slides four frames before the plant. Other task types ask you to author reference work yourself: a clean rig, a corrected animation pass, a shader breakdown, or a written critique that becomes training signal.

Some batches are pipeline-flavoured rather than artistic — evaluating whether generated Python for Maya or Blender actually runs, whether a Houdini node graph does what the prompt described, or whether a USD scene is assembled coherently. Others are pure craft judgment: silhouette readability, arcs, spacing, topology suitable for deformation, PBR values that survive a lighting change.

What the platform screens for

micro1's screen is AI-led and conversational, with follow-ups that go a level deeper than your first answer. It is checking that you have actually shipped 3D work — reel credits, shot counts, which software and which version, what you owned versus what the team owned — and that you can articulate craft judgment in words rather than only demonstrate it in a viewport. Expect to be pushed on specifics: name the rig control, name the render setting, describe the note a supervisor gave you and what you changed. Vague seniority claims collapse fast under those follow-ups.

Logistics

  • Fully remote, asynchronous, no fixed shift; work is claimed from queues that refresh irregularly
  • Most contributors take 10–25 hours per week; volume is not guaranteed and varies by specialisation
  • A reel or portfolio link is effectively mandatory — text-only applications rarely clear screening
  • Some tasks require you to have a working local install (Blender, Maya, Houdini, or equivalent) to reproduce or fix outputs
  • Rates observed in the $15–80/hr range; the upper band tracks technical specialisms like rigging, FX simulation, and pipeline tooling rather than generalist review