Tesla is paying up to $150K a year for people to watch and label videos
Tesla is hiring data labelers — people who manually tag objects in images and video — and offering total compensation of up to $150,000 a year. The catch: that figure includes base salary, bonuses, and equity. Per Glassdoor Tesla annotation salary benchmarks, the median base for this role sits closer to $67,000, with the 90th percentile around $107,000. Still, for work that requires no prior tech background, the package is unusually competitive.
The job
The role is based in Draper, Utah, and runs a standard 9-to-5:30 schedule with full benefits. Employees annotate footage from Tesla's vehicle fleet and its Optimus humanoid robots — drawing bounding boxes, flagging objects, and tagging scenarios so the neural networks behind Full Self-Driving and Optimus can learn to read the physical world. The Tesla official career posting is explicit: no previous experience in AI or data labeling required. Tesla provides on-the-job training.
Train the AI behind real-world robots
— Tesla Recruiting (@TeslaRecruiting) May 6, 2026
We’re hiring Data Labelers to annotate images and videos to train Optimus autonomous systems.
You’ll work with real production data directly impact how these systems learn and operate in the real world
Come join!… pic.twitter.com/ibqGimOBvV
The person driving this push is Duan Pengfei, Tesla's director of AI development, who has publicly stated his aim is to build the world's largest real-world data pipeline. Tesla has framed these positions not as back-office support but as foundational to Optimus becoming a viable product.
Why now
Optimus Gen 3 is moving toward limited production in 2026, according to the Optimus Gen 3 2026 roadmap. For a robot to function reliably outside a controlled factory, it needs to recognize thousands of real-world scenarios — and that recognition is built on annotated data. The more labeled footage the neural network ingests, the faster it generalizes.
Tesla's recruitment push for 1,000-plus human annotators is notable given the company simultaneously invests in automated labeling tools, including semi-automatic pipelines and 3D reconstruction models. Running both in parallel suggests the automated methods alone aren't yet producing data clean enough to train production-grade systems. Human review remains the quality checkpoint.
The automation paradox
There's an irony here that isn't lost on the industry: a company selling the promise of robots replacing human labor is paying humans good money to teach those robots how the world works. For job seekers without a coding background, these roles offer a genuine entry point into AI development — not as engineers, but as the people whose judgment shapes what the models actually learn.
Whether the $150K ceiling is achievable for most hires is a fair question. But even at median pay, Tesla is offering above-market rates for annotation work, signaling that data quality — not just data volume — is now a competitive priority.
