Introducing UpRock! DePin for AI - Share your idle internet and earn $UPT

Introducing UpRock - DePIN for the AI internet future

UpRock is an innovative DePIN network fueling the AI internet future that simplifies the process of auto-earning tokens, staking them, and redeeming them for rewards, making your journey in the web3 space as effortless as possible. Through our platform, users earn tokens by sharing their idle internet bandwidth and through staking, turning these tokens into real-world perks like flights and airtime minutes. With a growing community of 165,000 users and counting—gaining about 10,000 new users weekly and processing over 36 million transactions—we’re on a mission to make crypto assets more accessible and rewarding. Looking ahead, we aim to expand our services, deepen our collaboration with Solana, and partner with Jupiter to enhance the crypto experience for our users.

Check out our live network map

Describe your project in 5 sentences:
Imagine a wallet that’s more than just smart—it’s like your friendly sidekick. UpRock comes with AI that talks to you, helping you navigate the crypto world, finding deals, and earning for you while you’re off doing other things. The platform doesn’t just keep your crypto safe; it utilizes the depin network around the clock to turn your tokens into real-world perks, like cheaper flights and airtime minutes. It breaks down complex crypto concepts with help from its AI assistant, making every step of your crypto journey both friendly and straightforward. Our aim is to merge financial growth with real-life rewards on a platform that’s secure, easy to use, and genuinely cool to have.

Tell us about your team:

  • Team Members: The co-founders are a team of serial entrepreneurs with two acquisitions under their belt: The most recent Tenta Crypto Browser, the first fully encrypted mobile browser with built-in VPN was acquired by Avast, a publicly traded cybersecurity company.

  • Company Social Handles


Jupiter Ecosystem and You:

Our main pitch revolves around enhancing the liquidity and accessibility of our $UPT token within the Solana ecosystem through Jupiter’s exchange. By collaborating, we aim to offer seamless swap options for users, fostering a more interconnected and efficient crypto economy. Our integration with Jupiter aligns with our goal to provide users with intuitive financial tools and services that empower them to manage and grow their crypto assets effortlessly.

Potential synergies between you and Jupiter users, DAO, holders, etc.:
The synergy lies in providing Jupiter’s users with innovative ways to earn and grow their crypto assets through UpRock’s auto-earning features and exclusive deals. Additionally, UpRock users will benefit from Jupiter’s liquidity and swap functionalities, creating a mutual value exchange that enriches both communities.

Approximate date for TGE:
TBD (target is mid-March)

Vision Category:

If your project were to succeed, how would it fundamentally change the web3 space?

If UpRock succeeds, it’s going to change how people see tokens. Instead of just digital money or an investment, tokens could actually get you real things, like cheaper flights or special deals on services. It’s about making crypto part of daily life, showing it has practical uses beyond trading and investing. This could get more people interested in and using blockchain for the real benefits it can bring.

We also think the use of AI for customer service is largely ignored in web3 today. We’re completely embracing AI to simplify complex crypto concepts and interactions, creating a smart wallet that talks to you like a friend. This could make web3 appealing to a broader audience, turning the intimidating world of blockchain into an accessible and engaging experience for everyone.

What would you say is your biggest challenge or obstacle as a project?
In the crypto world, grabbing attention tends to be the most important skill, but the key is to build a product people love. It’s easy to get caught up in trying to make a big, short-term splash with points or token rewards, but our real goal is to build something that adds real value to our customers’ lives in the long run. This means focusing on creating a genuinely useful and loved product, not just getting initial validation. Our challenge is to never forget that.

Another one of our challenges at UpRock is navigating the relatively unexplored intersection of AI and blockchain. This pioneering space is thrilling because it offers the potential to make crypto more accessible and user-friendly, transforming it into a more human-centric experience. However, the novelty of integrating AI into blockchain also means we’re charting new territory, which comes with its own set of difficulties.

Creating an AI agent that our customers trust enough to act on their behalf is a significant undertaking. Trust in the digital world, especially within the crypto sphere, is hard-earned. Building an AI that not only understands but anticipates user needs, while ensuring security and privacy, requires careful planning, development, and continuous improvement. We’re excited about the potential of this technology to revolutionize how people interact with crypto, but we’re also mindful of the challenges in creating something truly reliable and effective in this new frontier.

What advice would you give to another team launching a token in web3?
When you launch a token, start by tackling a real problem your token can help solve. Think mobile-first to reach as many people as possible. Make sure your token is more than just an user acquisition strategy; it should be something that people actually want to use because it makes a difference.

What is something most tokens get wrong, and what steps are you taking to ensure that you won’t experience these pitfalls?

One common pitfall in the crypto space is the reliance on arbitrary time-based unlocks for tokens, which don’t align with a project’s real-world progress or success. Many tokens also miss having a solid utility or sustainable model for value creation. At UpRock, we’re addressing these issues head-on.

  1. Deflationary Transactions: We’re implementing deflationary actions within the UpRock platform to ensure $UPT’s scarcity and value increase as the platform grows.
  2. Active Tokenomics: We’re dedicated to making $UPT more than just a tradeable asset; it’s a crucial part of our ecosystem, used actively across our platform for various services and rewards, ensuring it circulates and retains value.
  3. Revenue Sharing: Our platform’s revenue goes back to our community, including liquidity providers and loyal users, with part of it used for buy-backs to redistribute $UPT, all managed transparently on-chain.

By aligning token unlocks with actual milestones, promoting in-platform circulation of $UPT, and sharing revenue with our users, we’re creating a token with genuine, sustainable value.

Update! - Attention iOS users

Sneak peak of the UpRock iOS app is near complete!



Socials links:



Excited to connect with the Jupiter community. Any comment or question is appreciated! :pray:


Looks like a great product in the current market given where technology is. what is the underlying infrastructure behind the AI? I think what’s more important is building the right infrastructure standards, as opposed to the actual end user wallet experience - at least in the current state of AI wealth management. More design color there would be powerful.


I feel confused after reading your post. So

  • What will you do with our ‘idle internet bandwidth’? How do you compare Uprock with Grass (

Thanks for the question!

GetGrass is great too and similar as far as building a specialized depin network for AI. However, UpRock stands apart for the following:

  • Wider Token Utility: UpRock emphasizes the broader utility of tokens through a rewards program that offers tangible benefits, such as exclusive deals and the option to swap tokens for real-world assets like travel or airtime minutes.

  • Aggregated Earnings: Instead of downloading dozens of apps, UpRock simplifies the process by combining various DePIN and earning opportunities into one platform. Later today we will announce one of our first partnerships!

  • AI-Powered UX: Enhances user experience with AI, making crypto more accessible and intuitive.

  • Leveraging Community Purchase Power: We’re building the best NFT-based rewards program. Once you get to millions of installs, we can leverage the community strength to secure better deals, enriching token value.


Do you have NFT collection as of now? or it’s on roadmap?


I’m intrigued…DePin is supposed to be the next big thing. I’m curious about a couple things, though. So this is an idle bandwidth play? but an AI assistant as well? The assistant can do web crawling, looking for deals. How is this going to scale, and what will it look like in 2 years? How many active users do you have on your app? What’s its current growth rate, and what will that look like with a JUP launch v without?
This looks like a very ambitious project. I’m curious to see how the execution goes.


Yeah we believe the DePIN network needs to go beyond just earning tokens and “cashing out” by swapping. A healthy depin network should also have a compelling rewards program where you can redeem and spend tokens within the network for real world goods and services. A reward program enhanced by AI insights to find the best deals online using the same network, enhancing token value. This program creates a strong flywheel effect where the platform generates additional commissions and fees for selling these deals, which is then shared with the network to provide more liquidity for data contributors.

Data crawling is very difficult to do well, but AI unlocks new methods of doing this more efficiently. As anyone who’s tried AI knows its mindblowing at pattern recognition, analyzing massive amounts of data, etc. which means the quality of the data products we sell will accelerate with AI. In two years, we expect to have millions of installs. That’s real browsers, real device peers that cant be easily censored or blocked, allowing more open and free flow of information in an AI internet future. Today we’re just scratching the surface with custom AI agents, but soon, people will trust them to perform tasks 24/7 and that requires agents to be able to freely collaborate and access the internet without being throttled or censored. We believe our data APIs will be a preferred alternative to the big tech censored versions.

Currently, we’re proud to have a growing community with 165k+ installs, adding roughly 10,000 new installs each week. Building large-scale mobile platforms is our skillset. When we launched MiKandi, because it was adult content, we could not get distribution through the Play Store or any mainstream outlet. We bootstrapped it to 15 million users on a shoestring budget before it was acquired. Mobile-first strategies are in our blood now. This momentum we have so far with UpRock is crucial as we look towards earning the support of the DAO to launch with Jupiter, which we anticipate will significantly accelerate our growth rate. However, if for any reason that does not happen, we will launch our IDO with our community and plan to list with as many partners that appreciate what we’re building.


Excited to share our first major partnership with!
@KoiiNetwork’s bustling marketplace of compute buyers & 15k+ desktop nodes meets UpRock’s growing 165K+ mobile network. Together, we set the stage for real revenue-generating opportunities in DePIN, meaning bigger rewards for everyone involved. LFG!

Learn more:


Thanks for the AI question! I shared a high level overview below of how we think about architecting our platform with AI, but if we missed the question, please let us know!

Starting with the tiiuae/Falcon-7B-Instruct model utilizing Hugging Face’s text generation inference toolkit for seamless operation and service delivery, we’re setting the stage for AI that engages in natural, human-like conversations. This model is just the beginning. We’re moving towards a network of specialized AI models, each expert in tasks like seat selection or reservations, working in concert to enhance user experiences. This network isn’t just about smart operations; it’s about creating detailed customer “artifacts” - rich data sets of attributes that the AI learns about the customer that is based on interactions, preferences, and behaviors tied to a preferable outcome. These artifacts enable us to tailor experiences and rewards, incentivizing user engagement through tokens.

We continuously train the model with our proprietary datasets to stay relevant and effective. This multi-model strategy enhances our platform’s capability to offer personalized, intuitive experiences. Our commitment to open standards and interoperability ensures our system is future-proof and integrates seamlessly with the broader web3 ecosystem.


Yes, we’re building an NFT-based rewards program. Where members can upgrade for more perks and benefits for interacting with the platform such as contributing to the network, doing transaction, staking, referring friends, interacting with the AI agent. Because Solana just released new token extensions and the new WNS standard was recently announced, we’re currently planning out what features would make a kickass reward program.


I’m pleased to share our user acquisition progress since being listed as a candidate on the LFG Launchpad.

While our average weekly user onboarding typically hovers around 10,000, on the 16th and 17th, we successfully onboarded 34,745 new users to UpRock, bringing our total user count to an impressive 201,000!

Let’s call it the #JupCandle moment because it truly is!

We are eagerly looking forward to pitching our startup to the community in the upcoming days.


What is the ratio of uprock points?


how can we see the uprock roadmap and tokenomics?


Hey there! Actually we’re not doing points, you’re earning tokens. They are just locked during beta phase and once we go mainnet, they will start unlocking.

We prefer tokens vs points because it keeps the team honest and committed (easy to cancel points program yanno), which adds healthy product development pressure on us to launch mainnet asap. Thanks for asking!


Moved here from Discord chat for space and time to ask and receive reply:

What security measures are in place to ensure the ‘unused bandwidth’ is being safely used only, with enough capacity built in for the home user? Also, there have been mixed reviews on a similar project, TouchGrass, can you address these? What program is monitoring what and are user passwords, surfing history etc protected? As it searches e.g. airline ticket prices, can you guarantee there will be no blocks from Cloudflare etc? Many thanks in advance.


Personal Privacy

Our app does not integrate with the end users’ browsers at all. This means we neither know nor could know what sites they visit, where they browse or anything else about their personal data. We are not collecting our end users’ data. Their devices are only being used as “workers” in our distributed network, the decisions about which sites to load and from which devices are driven by the data projects, meaning that an end user won’t inadvertently steer or influence this process in any way, and thus cannot leak data to it, even accidentally.

TouchGrass (GetGrass?)

Assuming that you mean Grass (getgrass), there’s actually a lot we think this project is doing right, and it seems like a hardworking and earnest team. That is to say, in the absence of them having done something shady (which we’re not aware of), we don’t want to throw shade.

We do think UpRock is superior to Grass, however. The main reasons for this are the nature of how work is distributed, the fact that UpRock works natively on mobile and the fact that UpRock is native app. In UpRock work providers never communicate directly with work producers, meaning that there is little to nothing that can leak about the user.

UpRock is also a native citizen on mobile. By far and away the most valuable and numerous users are mobile device users, and UpRock can address all of these. As a browser extension, Grass will always be blocked or severely limited on mobile devices.

UpRock cannot access the users’ browsers because it doesn’t run in the browser. Even if the Grass team never makes a dubious decision with their designs, the fact that it’s a browser extension means that a supply chain attack against the app will have the capability of injecting malware into the users’ browsers where their personal passwords, browsing data, etc live. In comparison, the UpRock app is designed to run with minimal permissions as a standalone app. Therefore even in a worst case scenario we aren’t in a position to be inside the users’ browsers. We are also not dependent on the user having Chrome running, and can run as a super lightweight background process.

Building native apps is more work than browser extensions, but the performance, security and reliability are worth “doing things right”.

Security Measure

We do not reveal our end users’ devices directly to data ingestion tasks, rather the devices run jobs which produce a result. Some of these jobs are interactive, but the end users’ devices are not hosting open ports and the data collection tasks do not communicate directly with the end user’s devices, but rather submit jobs to the coordinator nodes. These nodes then process these jobs with parameters specifying which sorts of devices (speed requirement, geo location, network type, etc) can fulfill these jobs, before returning the job result to the data task.

In terms of bandwidth usage, we have controls (and are developing more) to allow users to limit when tasks run, how much bandwidth they may use, etc. We expect that most users will want more bandwidth consumed on their devices, because this obviously provides them with bigger rewards, but if a user has some hard cap, or simply doesn’t want it running at certain times, they can control this as they like (and we will have more fine grained controls in the future, such as only accept tasks that pay more than some value, for certain types, etc).


We cannot totally guarantee that no process will ever hit a captcha (although with a DePin network this is much rarer). However, CAPTCHAs present another opportunity to reward our members. If tasks are blocked on CAPTCHAs we can allow our users to earn extra rewards for solving these CAPTCHAs. In fact, this is merely another way that users can increase their earnings. In fact, the user who solves the CAPTCHA doesn’t even need to be the same user whos device encountered the CAPTCHA, rather we can forward the CAPTCHA to a user who’s expressed an interest in earning for solving CAPTCHAs, and then forward the completed result back to the device that actually is doing the crawling (it’s a bit complicated, since we have to forward the solution not the result, but this encompasses the main ideas).


Our thoughts on the integration of AI & DePiN and the future possibilities

In our Discord discussions, there were questions regarding the evolution of our AI operations from initial, simpler tasks to more complicated functions as our product develops over time. Also, questions on AI’s integration with DePIN, as well as our potential impact on the broader adoption of Web3. @meow recommended I start posting responses in both places, so here we go!

First, I want to emphasize that we think a robust depin network shouldn’t just be about earning tokens and cashing out, but also have a super compelling reward/token redemption program, where you can get even more value from earned tokens for staking or spending them on exclusive deals within the system. That provides more commissions and fees for the platform, which is shared with the network. Imagine in a few years, we will have the collective power of millions of wallets. That means negotiating awesome deals for the network, for example bulk buying airtime minutes for our customers in Nigeria, so they can get discounts on charging their phone with tokens (even stake tokens to earn phone minutes). Planning ahead now means we will have the infrastructure to deliver a world class reward program.

Secondly, I believe all depin projects focus on their own token. However, just like Jupiter aggregates all liquidity sources, we can aggregate all depin sources for token earnings, since fundamentally they all demand the same resources (unique mobile IPs, bandwidth, some compute). We think it’s great value to offer users many tokens to earn, including Grass if they’d like to partner too.

Effortless Access

So first a crypto wallet that offers a conversational AI interface is going to be amazing, making it easier and more fun for users to get information about the company, product info and perform transactions. Right now, only sophisticated users are using crypto (yeah, we know there’s a range of sophistication, but compared to the “person on the street”, crypto users are sophisticated). And that’s great, it produces this great community. But, for crypto to reach its destiny, it needs everyone using it. People who think that Google is the web. People who can’t tell you the difference between DePin and a Pen. Yet there is a fundamental, irreducible complexity to crypto, therefore this ease of access cannot be solved by better UIs alone. Rather people will need a trusted guide to show them the way, answer their questions and provide sound guidance in order to bring more people to crypto and make it easier to use.

Complex Operation

Current AI chat apps have lowish memory limits issues because memory size explodes, making it expensive to process as the chat thread gets longer. Agents needs to be augmented with relevant data resources to present a brand or situation correctly. Furthermore, the agents ensemble needs to know when to invoke such augmentation, but also must be designed to smartly manage memory, distinguishing between critical and non-critical info to maintain essential customer details. All this, while prioritizing user privacy and not sharing customer data with third parties, ensuring that each user’s individual data remains private to them and doesn’t get mixed with public data.

One of the standout features we’re working on is the AI’s ability to identify and remember customer preferences that we call “customer artifacts,” which are key insights gained from interactions that have the propensity to lead to a favorable customer outcome. For example, (this is oversimplifying it) but if a customer frequently asks about Solana hackathons, the AI recognizes this pattern and labels it as a “customer artifact.” This particular artifact signifies that the “customer has a high propensity to attend Solana hackathons” and now the AI, with user permission, can proactively search deals 24/7 using the depin network!

AI find Patterns

With a significant DePin network grabbing data, it’s hard to know what’s noise and what’s an emerging trend. Whether it’s world events, changing preferences or just the latest online buzz, taking a bunch of disparate signals and aggregating them into a topic is a tedious and time consuming process. By the time some second rate site writes an article of “11 hottest trends”, an observant person will doubtless have already seen several of these. But identifying those trends as they emerge, across a diverse range of topics and disciplines is a borderline impossible problem without AI.

DePin Enhances AI

The other aspect is that AI will be supercharged with access to a DePin network to grow and learn about what’s going on around the world. Just like a curious toddler, models need exposure to tons of information and experiences in order to learn. Obviously, they cannot literally experience things, but they can learn from the experiences of others which collectively form the deluge of the web and social media. Giving the AI systems access to real time web data will not only allow them to answer questions more accurately, but fundamentally make it possible to train and specialize them on the myriad of human tasks and experiences which are shared online.

Where we’re going we need DePin & AI

In the future, the depin network can ingest more on-chain data as well. Which democratizes access to blockchain insights. Nobody needs to know a line of code in order to ask an agent “which top twenty wallets bought this NFT collection, JUP and WEN in the past 24 hours”. Lastly, if we’re going to build a depin network for AI, we would be less likely to succeed if we don’t know how AI agents or models work. Building our AI customer service agent requires us to dog-food our own network which will help us better win over data customers.


I signed up for a wallet to see what you have going on, fairly easy user on-boarding experience. My challenge however was what security do you have to protect the keys of each wallet?

Why the requirement to hold those keys or the wallet? Could this not be implemented similar to asset dash where you prove ownership to wallets through signing?


I wonder if this would help me clear or make real use of the mess in my wallet. There are a lot lot of tokens I simply avoid touching because I don’t know what they are.