Turing's 2024 AI pivot

This jump shows that Turing stopped behaving like a staffing marketplace and started behaving like an AI infrastructure vendor. In the old model, revenue rose one developer seat at a time. In the new model, frontier labs could buy large programs for coding data, model evals, and post training work, which let Turing scale spend much faster off the same supply network of vetted engineers and experts.1 2 3 4 5

• The numbers map the pivot clearly. Turing moved from about $120M in annualized revenue at the end of 2023 to $300M at the end of 2024, then $350M at the end of 2025. Growth slowed after the breakout year, which suggests 2024 was the step change when lab demand hit all at once.1 2
• What labs buy from Turing is not generic labeling. Turing sells pools of engineers and domain experts who write code, create hard test cases, generate training data, and score model behavior in workflows tied to reasoning, coding, agents, STEM, and multimodality. That work carries bigger budgets than placing one remote engineer at one company.3 4 5
• Turing now sits in a crowded human intelligence layer beside players like Scale AI, Surge, Mercor, micro1, Invisible, Handshake, and AfterQuery. The difference is that Turing came from remote developer staffing, so it already had a large technical talent base to redirect into AI lab work once that market opened.2 6 7 8

The next phase is likely less about hypergrowth from one time lab spending spikes, and more about turning this services base into a repeatable AGI platform. That means deeper eval products, more agent environments, and more enterprise AI work built from the same labor and data engine that powered the 2024 surge.3 5 9