Surge's Scale Advantage Over Micro1

This expansion shows that premium RLHF is becoming an operations scale business, not just a talent matching business. The winner is increasingly the company that can recruit thousands of workers, route each task to the right reviewer, catch bad labels automatically, and keep output flowing across time zones. That favors Surge, which pairs a much larger contractor base with software for live quality scoring, while Micro1 stays stronger in the narrowest, highest judgment expert categories.

• Micro1 is built like a selective marketplace. It screens thousands of candidates, accepts about 1 percent, and sells AI labs fast access to PhDs and senior engineers for tasks in medicine, law, physics, and coding. That model produces premium pricing, but it is harder to scale into very large annotation programs without broad labor supply.
  1 sacra
• Surge is built for bigger throughput. It reportedly had about 50,000 expert contractors and $1.2B of 2024 revenue, lets ML teams create tasks through a web app or SDK, and uses dashboards that track agreement scores, trust ratings, and automatic relabeling. New hubs and 5,000 more annotators deepen that advantage in volume and turnaround time.
  2 sacra
• The wider market is validating both models. Invisible scaled RLHF with a 3,000 plus person workforce and reached $134M of 2024 revenue, while Handshake used its academic network to reach $80M annualized AI labeling revenue by August 2025. That suggests frontier labs are buying from both full stack labor platforms and specialist expert networks, depending on task complexity.
  3 sacra 4 sacra

Going forward, the market is likely to split more clearly. Broad platforms like Surge will keep winning large, always on pipelines for model evaluation, red teaming, and multilingual review. Micro1's path is to own the highest value edge cases, where a lab needs a radiologist, tax lawyer, or senior engineer rather than a large managed workforce.