Mercor raises $750 million Series D at a $20 billion valuation to test enterprise AI agents
Mercor announced a $750 million Series D at a $20 billion valuation on 10 November 2026, doubling its value in 13 months to sell AI agent evaluation to enterprises.
By Emre A.Published
Tech AI LatestMercor, the San Francisco company that supplies human experts to train and evaluate AI models, announced a $750 million Series D at a $20 billion valuation on 10 November 2026. The price doubles the $10 billion valuation of its October 2025 Series C. Investors include Benchmark, General Catalyst, NVIDIA and Salesforce; no lead was named. The money goes to evaluation and post-training tools for enterprises. The round includes an employee tender offer of undisclosed size.
Key facts
- Announced: 10 November 2026 (Mercor, press release via Business Wire)
- Round: $750 million Series D at a $20 billion valuation, including an employee tender offer
- Previous round: $350 million Series C at $10 billion in October 2025
- Investors named: Benchmark, General Catalyst, NVIDIA, Salesforce, Sapphire Ventures and Zoom, among others; no lead named
- Network: more than 5 million experts, company figure
- Not disclosed: split between new capital and employee share sales, revenue in the release
What Mercor announced
Mercor announced on 10 November 2026 that it has raised $750 million in a Series D round at a $20 billion valuation. In a blog post titled "Organizing human intelligence for the entire economy" and a press release distributed by Business Wire, the company says the money will expand its evaluation and post-training infrastructure from frontier AI labs to enterprises across the economy.
The investors named include Benchmark, General Catalyst, NVIDIA, Salesforce, Sapphire Ventures and Zoom, along with Felicis, StepStone Group and XTX Ventures, according to the blog post. Mercor did not name a lead investor. The round includes a tender offer for employee shares, and the company has not said how much of the $750 million is new capital, as SiliconANGLE and other outlets note.
The valuation doubles the $10 billion Mercor reached with a $350 million Series C in October 2025, which Felicis led.
From frontier labs to enterprise agents
Mercor built its business by recruiting domain experts, such as doctors, lawyers, bankers and software engineers, and paying them to write training data and grade model outputs for AI labs. The company says its network now counts more than five million experts. Co-founder and co-chief executive Brendan Foody argues in the release that companies deploying AI agents will need to train and evaluate them on their own work, the same way the labs do with their models.
The enterprise product is custom evaluations: tests built from a company's real workflows that measure whether an agent meets that company's standards before it is put into production. Mercor says these build on APEX, its public benchmark series for economically valuable work. In July 2026 it acquired Deeptune, a company that spent two years rebuilding hundreds of enterprise applications as training environments where agents can practise; no price was disclosed. Coverage of the release, including Verdict, says the funds also pay for more such environments and new offices in New York and London.
Why the round matters
Evaluation has become one of the busiest corners of AI funding this autumn. Arena raised a $200 million Series B in October to grow its public model leaderboards, and Mercor's round is several times larger, for a company whose product is measuring and improving models rather than building them. The pitch to enterprises rests on a gap that model vendors have not closed: benchmarks published by OpenAI, Google or Anthropic describe general ability, not whether an agent can handle one insurer's claims process or one bank's reconciliation rules.
The investor list also shows where the demand comes from. NVIDIA and Salesforce both sell to companies building agents, and Salesforce runs its own agent platform. Mercor has said its revenue run rate passed $2 billion earlier this year; the funding release itself gives no revenue figure. Coverage of the round notes that most of that revenue has come from a small number of AI labs, so the enterprise push is also a way to reduce dependence on a few large customers.
What is not disclosed
Mercor has not said how the $750 million divides between new capital and the employee tender, what stake investors received, or who led the round. It has not named enterprise customers for the evaluation product or given prices. The next signals to watch are the first named enterprise deployments and whether the labs that make up most of its business keep buying expert data as they build more of it in house. The round is listed in the AI funding tracker, and more deals are on the AI business hub.
What changed
| Before | Mercor was valued at $10 billion after a $350 million Series C in October 2025 and sold mainly to frontier AI labs |
|---|---|
| After | Mercor is valued at $20 billion after a $750 million Series D and is selling custom agent evaluations and training environments to enterprises |
| Who it affects | Enterprises deploying AI agents, AI labs that buy expert training data, the experts in Mercor's network, rival evaluation and data companies |
Questions people ask
- How much did Mercor raise?
- $750 million in a Series D at a $20 billion valuation, announced on 10 November 2026. The amount includes an employee tender offer, and Mercor has not said how much is new capital.
- Who invested in Mercor's Series D?
- Investors named by Mercor include Benchmark, General Catalyst, NVIDIA, Salesforce, Sapphire Ventures, Zoom, Felicis, StepStone Group and XTX Ventures. No lead investor was named.
- What does Mercor do?
- It pays a network of more than five million domain experts, by its own count, to create training data and evaluate AI models, and it now builds custom evaluations and training environments for companies deploying AI agents.
Sources
- Organizing human intelligence for the entire economy (Mercor, 10 Nov 2026)
- Mercor Raises $750 Million Series D at $20 Billion Valuation to Help Enterprises Own Their Intelligence (Mercor (Business Wire), 10 Nov 2026)
- Mercor raises $750M at $20B valuation to take AI agent evaluation to enterprises (SiliconANGLE, 10 Nov 2026)
- Mercor raises $750m to expand AI evaluation tools for enterprises (Verdict, 11 Nov 2026)
- Mercor Raises $750M Series D at $20B Valuation to Expand AI Training (Unite.AI, 10 Nov 2026)
Reference pages
- Mercor company
- NVIDIA company
- Salesforce company
Spotted an error? See the corrections policy and tell the editor. Facts in this story come from the sources listed above; vendor claims are attributed.
Comments
Loading comments.





