This robotics data company once aimed to start with Crypto points.

Written by: KarenZ, Foresight News

A project that once attracted users to contribute data through Crypto-style points has now gained the support of tech giants and investment institutions such as Sequoia Capital, Nvidia, and Microsoft, and is expected to generate $300 million in revenue by the end of the year.

This project is Mecka AI.

On October 7, Mecka AI announced the completion of a $60 million Series B funding round, led by Sequoia Capital, with new investors including Nvidia, Microsoft's M12, Qualcomm Ventures, and Samsung participating, while Framework Ventures, Kindred Ventures, and incubator VC Neo continued their support.

The company also disclosed that its annualized revenue run rate as of June 2026 has exceeded $100 million, with expectations to reach $300 million by the end of the year. It should be noted that this is the company's announced annualized revenue run rate, not the actual annual revenue achieved.

For the Crypto industry, Mecka is not a completely unfamiliar AI company. It not only has investors from the crypto industry, but the founding team also has a Crypto background. More notably, Mecka previously launched a points program colored by the crypto industry, which once sparked speculation among users about its future token issuance.

However, nowadays, Mecka's official website and funding announcements hardly emphasize Web3, but instead position itself as a Physical AI data infrastructure company.

What changes has Mecka undergone from Crypto points to robotics data business?

What is the background of Mecka?

Mecka's four main co-founders are: Josh Gao (CEO), Jason Chong (CTO), Duy Nguyen (COO), and Mogen Cheng (CPO).

Interestingly, none of the four had a professional background in robotics research initially. CEO Josh Gao co-founded a catering software startup with Mogen Cheng and sold the business in 2023.

Co-founder and CTO Jason Chong is an old face in the Crypto industry; he founded the crypto payment startup Utopia Labs, which was later acquired by Coinbase. According to Fortune magazine, operations manager Duy Nguyen has experience in the sneaker resale business, where he made millions of dollars through reselling sneakers.

Josh Gao acknowledged in a Fortune interview that team members do not have backgrounds in the robotics industry. They discovered a problem through reading research papers, visiting robotics labs, and communicating with researchers: the robotics industry has increasingly advanced hardware and models, yet lacks large-scale, real, and sufficiently rich human motion data. The team thus set its direction: instead of directly manufacturing robots, they would provide the data needed for robots to learn about the real world.

Regarding the founding date, public reports show some discrepancies. TechCrunch traces the founding of Mecka back to 2024, while Fortune states that the four founders decided to formally launch Mecka AI in 2025 after accumulating substantial data in the early stages. Therefore, 2024 can be understood as the early exploratory stage of the project, while 2025 is the company’s launch phase as stated in reports.

In terms of funding, Mecka publicly completed an $8 million seed round in August 2025, led by Neo.

In June 2026, the company announced it had completed a total of $60 million in funding (consisting of $25 million in Series A and $35 million in subsequent investments), led by Framework Ventures, with participation from Menlo Ventures, SV Angel, Kindred Ventures, and others.

Four months later, Mecka secured another $60 million in Series B funding, with the investor lineup expanding further to include Sequoia Capital, Nvidia, Microsoft's M12, and others.

From Framework Ventures to Sequoia Capital, and to strategic investors like Nvidia and Microsoft, Mecka's funding lineup is no longer limited to the Crypto circle.

What is Mecka?

AI can write articles and generate code, but to teach robots to serve coffee, fold clothes, and repair cars still requires a lot of training.

The reason is simple: language models can access vast amounts of text from the internet, but robots need to understand actions, forces, contacts, and spatial relationships in the real world. This information cannot be obtained directly from webpages.

What Mecka does is convert human daily activities into training data that robots can use.

Its operating process mainly consists of three steps.

First is data collection. Mecka organizes personnel to use mobile phones or self-developed multi-sensor devices to record daily activities such as cooking, organizing items, and car repairs.

Then, data processing. The company utilizes computer vision, multimodal models, and other technologies to process the raw videos and sensor data, extracting action trajectories, hand postures, three-dimensional spatial information, and more.

Finally, data delivery. Mecka provides the processed data to robotics labs, AI model teams, and enterprises for model training, evaluation, and actual robot deployment.

Mecka's core business model is B2B data and related services.

In August 2025, according to Upstarts, Mecka disclosed that it had signed paid contracts with some Fortune 100 companies, listing robotics company 1X Technologies as an early partner.

Currently, the company claims to have provided services to several leading robotics labs and several of the top seven tech giants in the United States, but has not disclosed a complete client list or specific contract amounts.

However, Mecka is no longer satisfied with just selling data.

In the latest funding announcement, the company also emphasized its robotic commercial deployment business, including hardware access, on-site data collection, model post-training, system integration, and subsequent operations maintenance.

This means Mecka hopes to evolve from a training data supplier to a service provider for enterprise robot deployment.

The value of this model lies in the fact that Mecka can not only sell one-off data but may also establish long-term cooperative relationships with customers through deployment and ongoing operations.

Has Mecka downplayed Web3?

Compared to its current positioning as a physical AI, Mecka's early Crypto flavor was much more pronounced.

In 2024, Mecka launched a points program inspired by the crypto industry. The early points even covered social tasks, serving a user growth function.

Regarding this experience, according to Upstarts, Mecka CEO Josh Gao provided an explanation. This model, while helpful for garnering attention in the project's early stages, also brought trust issues for the company, especially when cooperating with large enterprises. CTO Jason Chong believes that if a company wishes to collect millions of hours of variable-quality video data, a token mechanism might work; but if such a scale of collection is not needed, it may also become a distraction.

Subsequently, Mecka turned to gamifying daily tasks, allowing users to contribute data through activities like making beds. For custom tasks for enterprises, the company provides specialized equipment to selected data collection personnel to obtain higher quality data.

So, does Mecka still issue points to users now? This remains unconfirmed.

In February 2026, Mecka launched a new iOS app on the Apple App Store, which the company describes as a first-person data recording tool aimed at internal robotics research labs, primarily for data recording, uploading, and task management, without publicly mentioning any points, tokens, or airdrop mechanisms.

I downloaded the app, but upon entering, I needed to fill in an Access Code, so I couldn't view internal tasks and reward rules for now, making it impossible to determine whether current data providers are receiving points, cash, or other forms of compensation.

When the data business is established, is a token still needed?

Looking back at Mecka's development, the most noteworthy aspect is not whether it will issue a token, but how its data collection model gradually aligns with enterprise needs.

The early points program addressed the cold start problem: how to attract more ordinary users to participate and provide a lot of real-world data for robots?

However, the needs of commercial clients are more complex.

For robotics labs, the quantity of videos is not the only metric. Accuracy of data, completeness of actions, correspondence to specific training tasks, and stability of collection quality are also equally important.

Therefore, an open-ended points incentive and specialized paid collection are not completely the same business model.

The former can more easily attract a large number of participants, while the latter is conducive to managing the collection process around specific client needs. Both methods could coexist, and it's not simple to determine which one is necessarily more effective.

Mecka's early attempt at Crypto points followed by a greater emphasis on enterprise custom data and commercial deployment can be seen as a gradual focus of its business model.

And over a $100 million annualized revenue run rate at least indicates that the company claims its current business has formed a substantial revenue base. However, due to the lack of complete financial disclosure, outsiders still cannot judge its gross margin, profitability, and growth sustainability.

Mecka's development also raises a thought-provoking question for the Crypto industry: If a project can ultimately rely on traditional commercial revenue to continuously pay data contributors and meet enterprises' data procurement needs, is a token still a necessity?

For Mecka, this question has not yet had a final answer.

But it is certain that the points program that once attracted Crypto users’ attention is no longer the focus of its external promotion. Instead, the focus has shifted to robotics training data, enterprise clients, commercial deployment, and the ever-increasing revenue scale.