Updated 8 Nov 2026
The best AI data companies in 2026, ranked by the money on record.
The largest companies selling AI developers what their models learn from are Mercor ($2B, run rate, Jun 2026), Surge AI ($1.2B, FY2024), Handshake AI ($1B, run rate, Aug 2026), micro1 ($1B, run rate, Oct 2026) and Scale AI ($870M, FY2024). Every company here is ranked by what it earns, then what it is valued at, then what it raised, and every figure links to the filing or article that states it. No company pays for a place.
Hold data these companies buy? AIData.Broker values it free and introduces it to every buyer that takes it, at no cost to you: the buyers pay us. Where it earns more, we run the sale ourselves, for a fee only out of a sale.
- No data upload
- About two minutes
- No email needed
The largest AI data companies by revenue
The ten with the most revenue on record, each bar linked to the record it comes from.
The latest money moves
The newest rounds, valuations and revenue figures across the register, each linked to its record.
- 5 Nov 2026Innodata reported $350M revenue (12 months to Sep 2026)
- 4 Nov 2026TaskUs reported $250M revenue (12 months to Sep 2026, AI Services)
- 28 Oct 2026Plato raised $16M
- 27 Oct 2026CVAT raised $2M
- 22 Oct 2026Maxinsights raised $30M
- 22 Oct 2026BioStack Platforms raised $9.5M
- 14 Oct 2026Labelbox raised $100M at a $1.3B valuation
- 12 Oct 2026micro1 was valued at $4B valuation
Every ranking
One list per kind of work, each answering the search people make for it, ranked by the same rule.
- Best AI data labs and human data companies
98 companies, 41 ranked
Mercor $2B, Surge AI $1.2B, Handshake AI $1B. They pay experts and contractors to write the tasks, demonstrations, rubrics and judgments AI labs train and test their models on.
- RL environment companies
91 companies, 36 ranked
Mercor $2B, Surge AI $1.2B, Handshake AI $1B. They build the simulated software, workplaces and tasks AI agents practise in, and the checks that score each attempt.
- Best data labeling companies
77 companies, 40 ranked
Handshake AI $1B, Scale AI $870M, Prolific $500M. They label, annotate and review data for AI teams, with their own workforce, a platform, or both.
- Best data labeling platforms
20 companies, 14 ranked
Scale AI $870M, Innodata $350M, Appen $249M. They sell the software a team labels, annotates and reviews its own training data in, whether or not they also label it for you.
- AI evaluation companies
137 companies, 60 ranked
Mercor $2B, Surge AI $1.2B, Handshake AI $1B. They build the benchmarks, test sets and expert reviews that measure what a model or an agent can do.
- Best AI training data providers
115 companies, 53 ranked
Mercor $2B, Surge AI $1.2B, micro1 $1B. They sell datasets, made to order or off the shelf, to the companies that train AI models.
- Coding data companies for AI
70 companies, 33 ranked
Mercor $2B, Surge AI $1.2B, Handshake AI $1B. They supply the code, programming tasks, software environments and reviews that coding models and agents learn from.
- Robotics training data companies
61 companies, 31 ranked
Scale AI $870M, Innodata $350M, TaskUs $250M. They record, label and simulate the physical world for robot learning: teleoperation, egocentric video, touch and motion.
- Medical AI training data companies
56 companies, 26 ranked
Mercor $2B, Surge AI $1.2B, micro1 $1B. They supply clinical records, medical images and the judgments of doctors and nurses for health AI.
- Speech and audio training data companies
38 companies, 20 ranked
Surge AI $1.2B, micro1 $1B, Innodata $350M. They record, transcribe and license speech and audio for voice models, in many languages and accents.
- Video and image training data companies
68 companies, 37 ranked
Surge AI $1.2B, Handshake AI $1B, Innodata $350M. They license, capture and label footage and pictures for vision, video and world models.
- Finance AI training data companies
54 companies, 25 ranked
Mercor $2B, Surge AI $1.2B, Handshake AI $1B. They supply financial documents, analyst tasks and expert judgments for models that work in banking, investing and accounting.
- Legal AI training data companies
32 companies, 17 ranked
Mercor $2B, Surge AI $1.2B, Handshake AI $1B. They supply legal documents, lawyers' tasks and expert reviews for models that work in law.
- AI red teaming and security data companies
23 companies, 10 ranked
Proximal $200M, Toloka $26.4M, Vals AI $400M valuation. They attack models and agents the way an adversary would, and sell the security tasks, exploits and evaluations that make them safer.
- Multilingual training data providers
34 companies, 17 ranked
Surge AI $1.2B, Prolific $500M, Innodata $350M. They supply text, speech and expert judgments in many languages, so models work beyond English.
- Synthetic training data companies
31 companies, 15 ranked
Mercor $2B, Appen $249M, Proximal $200M. They generate training data with models, simulators and code rather than record it, and sell it to the companies that train AI.
- Best AI data collection companies
58 companies, 30 ranked
Innodata $350M, TaskUs $250M, Appen $249M. They record new data from people and the real world for AI training: speech, video, motion, egocentric footage and the like.
- Computer-use agent data companies
37 companies, 17 ranked
Scale AI $870M, Snorkel AI $375M, Turing $300M. They record people using software and build replicas of real apps, so agents learn to click, type and finish work on a screen.
- Who pays the most for company data
Buyers by their published payouts
The AI data buyers that publish what they pay a company for its records, highest first, each in its own words.
- Data brokers for AI
Compared on their published terms
The brokers, marketplaces and wind-down services a company sells its data through: what each pays the company and when, in its own words.
- The market map
Every company on one page
Each company in a box for each thing it sells, the largest first, and the RL environment companies by the work their environments are for.
- Head to head
The comparisons people search
Scale AI vs Surge AI, Mercor vs Scale AI, Mercor vs micro1 and more, side by side on the same figures, and every ranked company against its competitors.
The whole register, ranked
Each company is ranked on the first of these it has on record: its revenue over twelve months (a run rate or a year as published, or its last four quarters added up; an AI segment where the rest of its business is not AI data), then its newest valuation, then the money it disclosed raising. No company pays for a place.
| # | Company | Ranked on | Measure | Valuation | Sells | Pays companies for data |
|---|---|---|---|---|---|---|
| 1 | Mercor | $2B | Revenue, run rate, Jun 2026 | $10B | RL environments, Datasets, Expert and human data, Evaluations and benchmarks | Through our introductions |
| 2 | Surge AI | $1.2B | Revenue, FY2024 | RL environments, Datasets, Expert and human data, Evaluations and benchmarks | ||
| 3 | Handshake AI | $1B | Revenue, run rate, Aug 2026 | $3.5B | RL environments, Evaluations and benchmarks, Expert and human data, Annotation | Through our introductions |
| 4 | micro1 | $1B | Revenue, run rate, Oct 2026 | $4B | RL environments, Evaluations and benchmarks, Expert and human data, Datasets | Through our introductions |
| 5 | Scale AI | $870M | Revenue, FY2024 | $29B | Labeling software, RL environments, Datasets, Expert and human data, Evaluations and benchmarks, Annotation, Simulation | Through our introductions |
| 6 | Prolific | $500M | Revenue, run rate, May 2026 | Evaluations and benchmarks, Expert and human data, Datasets, Annotation | ||
| 7 | Snorkel AI | $375M | Revenue, run rate, Sep 2026 | $3.5B | Datasets, RL environments, Expert and human data, Evaluations and benchmarks, Annotation | |
| 8 | Innodata | $350M | Revenue, 12 months to Sep 2026 | Labeling software, RL environments, Datasets, Expert and human data, Evaluations and benchmarks, Annotation | ||
| 9 | Turing | $300M | Revenue, run rate, Mar 2025 | $2.2B | RL environments, Datasets, Expert and human data, Evaluations and benchmarks | Through our introductions |
| 10 | TaskUs | $250M | Revenue, 12 months to Sep 2026 (AI Services) | Annotation, Expert and human data, Datasets | ||
| 11 | Appen | $249M | Revenue, 12 months to Jun 2026 | Labeling software, Datasets, Expert and human data, RL environments, Evaluations and benchmarks, Annotation | Its own programme | |
| 12 | Proximal | $200M | Revenue, run rate, Sep 2026 | $300M | Evaluations and benchmarks, Datasets | |
| 13 | Invisible Technologies | $134M | Revenue, FY2024 | $2B | RL environments, Datasets, Expert and human data, Evaluations and benchmarks, Annotation | |
| 14 | AfterQuery | $100M | Revenue, run rate, Apr 2026 | $3.2B | RL environments, Datasets, Expert and human data, Evaluations and benchmarks | Through our introductions |
| 15 | Arena | $100M | Revenue, run rate, Oct 2026 | $3.1B | Evaluations and benchmarks, Datasets | |
| 16 | Mecka | $100M | Revenue, run rate, Jun 2026 | $500M | Datasets, Expert and human data, Evaluations and benchmarks | |
| 17 | Wirestock | $40M | Revenue, run rate, May 2026 | Datasets, Expert and human data | Its own programme | |
| 18 | Toloka | $26.4M | Revenue, FY2024 | RL environments, Datasets, Expert and human data, Evaluations and benchmarks, Annotation | ||
| 19 | Eigent AI | $12M | Revenue, run rate, Oct 2026 | RL environments, Datasets, Evaluations and benchmarks | ||
| 20 | Applied Intuition | $15B | Valuation, 2025 | Simulation, RL environments, Evaluations and benchmarks, Annotation | ||
| 21 | Labelbox | $1.3B | Valuation, 2026 | RL environments, Expert and human data, Evaluations and benchmarks | Its own programme | |
| 22 | Mechanize | $500M | Valuation, 2026 | RL environments, Evaluations and benchmarks, Datasets, Simulation | ||
| 23 | Vals AI | $400M | Valuation, 2026 | Evaluations and benchmarks, Datasets | ||
| 24 | Kled | $100M | Valuation, 2025 | Datasets, Expert and human data, Evaluations and benchmarks, Annotation | Its own programme | |
| 25 | Lightwheel | $145M | Money raised, disclosed | Datasets, Expert and human data, Evaluations and benchmarks, Annotation, Simulation | ||
| 26 | Encord | $90M | Money raised, disclosed | Labeling software, Annotation, Expert and human data, Datasets, Evaluations and benchmarks | ||
| 27 | Patronus AI | $70M | Money raised, disclosed | RL environments, Datasets, Expert and human data, Evaluations and benchmarks, Simulation | Through our introductions | |
| 28 | Sama | $70M | Money raised, disclosed | Expert and human data, Annotation, Evaluations and benchmarks | ||
| 29 | XDOF | $70M | Money raised, disclosed | Evaluations and benchmarks, Datasets, Annotation | ||
| 30 | CloudFactory | $65M | Money raised, disclosed | Datasets, Annotation, Expert and human data, Evaluations and benchmarks | ||
| 31 | Defined.ai | $62.3M | Money raised, disclosed | Datasets, Expert and human data, Annotation, Evaluations and benchmarks | Its own programme | |
| 32 | Centific | $60M | Money raised, disclosed | RL environments, Datasets, Expert and human data, Evaluations and benchmarks, Annotation | Its own programme | |
| 33 | Roboflow | $60M | Money raised, disclosed | Labeling software | ||
| 34 | David AI | $50M | Money raised, disclosed | Datasets, Expert and human data, Evaluations and benchmarks | ||
| 35 | SuperAnnotate | $49.5M | Money raised, disclosed | Labeling software, Expert and human data, RL environments, Evaluations and benchmarks, Annotation, Datasets | Its own programme | |
| 36 | Deeptune | $43M | Money raised, disclosed | RL environments, Datasets | ||
| 37 | Voxel51 | $42.5M | Money raised, disclosed | Labeling software | ||
| 38 | Bespoke Labs | $40M | Money raised, disclosed | RL environments, Evaluations and benchmarks, Datasets, Simulation | ||
| 39 | Gray Swan AI | $40M | Money raised, disclosed | Evaluations and benchmarks, Datasets | ||
| 40 | Dataloop | $33M | Money raised, disclosed | Labeling software, Annotation | ||
| 41 | Kili Technology | $32M | Money raised, disclosed | Labeling software, Annotation, Expert and human data, Evaluations and benchmarks | ||
| 42 | Luel | $31.2M | Money raised, disclosed | Datasets, Expert and human data, Evaluations and benchmarks, Annotation | ||
| 43 | Halluminate | $30M | Money raised, disclosed | RL environments, Evaluations and benchmarks | Its own programme | |
| 44 | Maxinsights | $30M | Money raised, disclosed | Expert and human data, Evaluations and benchmarks, Annotation, Simulation | ||
| 45 | Moonlake AI | $28M | Money raised, disclosed | Evaluations and benchmarks, Simulation | ||
| 46 | Deccan AI | $25M | Money raised, disclosed | RL environments, Datasets, Expert and human data, Evaluations and benchmarks, Annotation | ||
| 47 | HumanSignal | $25M | Money raised, disclosed | Labeling software, Annotation, Evaluations and benchmarks | ||
| 48 | iMerit | $23.5M | Money raised, disclosed | Labeling software, Expert and human data, Evaluations and benchmarks, Annotation, Datasets | ||
| 49 | SimpleClosure | $19M | Money raised, disclosed | Datasets | Its own programme | |
| 50 | Taste Labs | $18.5M | Money raised, disclosed | Expert and human data | ||
| 51 | HUD | $16M | Money raised, disclosed | RL environments, Evaluations and benchmarks | ||
| 52 | Plato | $16M | Money raised, disclosed | RL environments | ||
| 53 | Preference Model | $16M | Money raised, disclosed | RL environments | ||
| 54 | Sapien | $15.5M | Money raised, disclosed | Evaluations and benchmarks, Annotation | ||
| 55 | Datacurve | $15M | Money raised, disclosed | RL environments, Datasets, Expert and human data, Evaluations and benchmarks, Annotation | Its own programme | |
| 56 | Midcentury | $15M | Money raised, disclosed | Datasets, Expert and human data, Simulation, Evaluations and benchmarks, Annotation | ||
| 57 | General Reasoning | $10.9M | Money raised, disclosed | RL environments, Evaluations and benchmarks | ||
| 58 | Chakra Labs | $10.1M | Money raised, disclosed | RL environments, Expert and human data, Evaluations and benchmarks, Datasets | ||
| 59 | Arga Labs | $10M | Money raised, disclosed | RL environments, Evaluations and benchmarks | ||
| 60 | BioStack Platforms | $9.5M | Money raised, disclosed | RL environments, Datasets, Evaluations and benchmarks, Simulation | ||
| 61 | Idler | $9M | Money raised, disclosed | RL environments, Datasets, Evaluations and benchmarks, Simulation | Its own programme | |
| 62 | Perle AI | $9M | Money raised, disclosed | Datasets, Expert and human data, Evaluations and benchmarks, Annotation | ||
| 63 | Veris AI | $8.5M | Money raised, disclosed | Evaluations and benchmarks, RL environments | ||
| 64 | Human Archive | $8.2M | Money raised, disclosed | Datasets, Expert and human data, Annotation | ||
| 65 | Runloop | $7M | Money raised, disclosed | Evaluations and benchmarks, Datasets, RL environments | ||
| 66 | clickworker | $6M | Money raised, disclosed | Datasets, Annotation | ||
| 67 | VisionLab AI | $6M | Money raised, disclosed | Datasets, Annotation, Expert and human data | ||
| 68 | Build AI | $5M | Money raised, disclosed | Datasets, Evaluations and benchmarks | ||
| 69 | Latch | $5M | Money raised, disclosed | Evaluations and benchmarks, Datasets, Annotation | ||
| 70 | Troveo | $4.5M | Money raised, disclosed | Datasets, Annotation | Its own programme | |
| 71 | Macrodata Labs | $4M | Money raised, disclosed | Annotation, Evaluations and benchmarks | ||
| 72 | Good Start Labs | $3.6M | Money raised, disclosed | RL environments, Datasets, Expert and human data, Evaluations and benchmarks, Annotation | ||
| 73 | Lightly | $3M | Money raised, disclosed | Labeling software | ||
| 74 | Gradient Health | $2.8M | Money raised, disclosed | Datasets | Says it buys data | |
| 75 | Poindexter Labs | $2.7M | Money raised, disclosed | Datasets, Expert and human data, Evaluations and benchmarks, Annotation | ||
| 76 | CVAT | $2M | Money raised, disclosed | Labeling software, Annotation | ||
| 77 | Quesma | $2M | Money raised, disclosed | RL environments, Evaluations and benchmarks | ||
| 78 | Segments.ai | $1.1M | Money raised, disclosed | Labeling software, Annotation |
- Who is on it.
- Every company on our vendor register: companies that say on their own pages that they sell AI developers training data, expert work, labeling, RL environments or evaluations, read word for word.
- Pays companies for data.
- "Through our introductions": it takes companies' data through AIData.Broker's introductions and pays us a referral fee when one goes ahead, which never moves its place. "Its own programme" and "Says it buys data" are its own published terms and words.
Short answers
- Which are the largest AI training data companies?
- By the money on record: Mercor, $2B (run rate, Jun 2026); Surge AI, $1.2B (FY2024); Handshake AI, $1B (run rate, Aug 2026). 19 of the 179 on the list have revenue on record; the others are ranked on their valuation or the money they raised.
- Is Scale AI the biggest of the AI training data companies?
- By valuation, yes: Scale AI was valued at $29B (Jun 2025). By revenue on record, Mercor is larger: $2B (run rate, Jun 2026), against $870M (FY2024) for Scale AI.
- Which AI training data companies buy data from companies?
- 31 of them take data from companies that hold it, on a published programme or in their own words: Mercor, Handshake AI, micro1, Scale AI, Turing, Appen, AfterQuery, Wirestock, Labelbox and Kled, and others in the table. A company holding such data can value it free and be introduced to every one whose rules take it.
- How is this list ranked?
- Each company is ranked on the first of these it has on record: its revenue over twelve months (a run rate or a year as published, or its last four quarters added up; an AI segment where the rest of its business is not AI data), then its newest valuation, then the money it disclosed raising. No company pays for a place.
Cite this
AIData.Broker, “The best AI data companies in 2026, ranked by the money on record”, reviewed 8 Nov 2026. https://aidata.broker/best
Free to reuse under CC BY 4.0 if you credit AIData.Broker and link to this page. Each row names the publication that stated it, and words quoted from a source keep that source's own terms.
Your is downloading.