Turing took an unusual road into the AI training data market. It spent its first years as a managed marketplace for remote software engineers, then discovered that the same vetted network was exactly what frontier labs needed for coding data, evaluations, and post-training work. The pivot reportedly tripled revenue to around 300 million dollars in 2024 while reaching profitability, with OpenAI, Google, Anthropic, and Meta among reported customers. A vendor serving every frontier name ends up on every shortlist, and shortlists produce comparison shopping. This guide covers who competes with Turing in each lane, and when the better answer is not an expert network at all.
What Turing is known for
Turing's asset is a developer network it reports at more than 4 million engineers and domain experts, built during its staffing years and screened by its vetting stack. On top of it sit three businesses: the original talent placement service, an AI lab services arm supplying coding, reasoning, STEM, multimodal, and agentic training data, and an enterprise arm deploying AI systems for corporations. For AI buyers, the middle one is what matters: expert engineers producing the demonstrations, evaluations, and code data that current models train on.
The reported numbers sketch the trajectory: roughly 120 million dollars of revenue in 2023, around 300 million in 2024 with profitability, and a 111 million dollar Series E in 2025 at a 2.2 billion dollar valuation. Founded in 2018 by Jonathan Siddharth, it is one of the longer-tenured names in a market full of two-year-old companies.
Teams look for alternatives for familiar reasons: diversifying vendors on mission-critical programs, wanting depth in domains beyond software engineering, preferring a pure-play data company over one balancing three business lines, or realizing the gap is training data that exists in the world rather than work experts produce to order.
Alternatives for coding and expert data
The direct competitors are the expert networks, each with a different engine. Mercor is the largest of the new generation, recruiting working professionals across industries and expanding into evals and RL environments. Micro1 vets applicants with its AI recruiter at high selectivity and runs lean. Handshake AI draws PhDs from its university career network, strongest in academic domains. Surge AI is the premium name for frontier-grade human feedback, and Scale AI runs the largest full-stack programs, with the Meta ownership caveat every buyer weighs; our guide to Surge AI alternatives maps that premium end. AfterQuery pairs expert data with published benchmarks, and Toloka and Prolific cover managed crowds and vetted participants.
For the human feedback lane in full, see our guide to RLHF data providers.
When the answer is data, not experts
Expert networks sell knowledge work produced to order, and Turing's lane is the purest version of it: code written, reviewed, and evaluated by engineers. That is the right purchase when your model lacks skill or judgment. It is the wrong purchase when your model lacks contact with the real world: video models need real motion, voice agents need natural conversation, world models need gameplay and first-person footage, agents need genuine workflow traces from real businesses. No network of engineers can write that data. It has to be captured from reality and licensed from whoever owns it, a different market covered in our guide to AI training data marketplaces.
The landscape
| Provider | Category | Best for |
|---|---|---|
| Turing | Expert data and services | Coding data, expert engineers, lab post-training work |
| Mercor | Expert marketplace | Domain-expert demonstrations, evals, environments |
| Micro1 | Expert recruiting | AI-vetted specialists, lean challenger economics |
| Handshake AI | Expert network | PhD-level reasoning data from a university pipeline |
| Surge AI | RLHF and evaluation | Frontier-grade preference data |
| Scale AI | Full-stack data services | Very large programs, Meta caveat applies |
| AfterQuery | Expert data and benchmarks | Benchmarked SFT and RLHF data |
| Toloka | Managed global crowds | Scaled human data with global reach |
| Prolific | Vetted participant pool | Representative human feedback and evals |
| Troveo | Licensed data marketplace | Rights-cleared video, audio, gameplay, and business data |
How to choose
Three axes separate the expert vendors. Domain center of gravity: Turing is deepest in software engineering, Handshake AI in academic domains, Mercor across professions. Focus: Turing balances staffing, lab services, and enterprise work, while the newer networks are pure-play training data companies, which cuts both ways between stability and concentration. And tenure versus momentum: Turing has been operating since 2018 and reports profitability, unusual among competitors burning capital to grow. Whoever you pick, confirm your program gets dedicated expert capacity rather than whatever the network has left over.
And if what your model is missing is the world itself rather than expert judgment about it, stop comparing networks: that budget belongs in licensed data.
Where Troveo fits
Troveo does not run an expert network. Troveo licenses the data engineers cannot write: real-world video, audio, gameplay, and business data from more than 7,000 rights holders, cleared for AI training with documentation per asset and delivered training-ready. The two purchases are complements, and serious model programs make both. Browse the catalog in Lens or talk to us about what your model is missing.