The Bittensor ecosystem is expanding fast — and no one tracks its pulse better than the TAO community itself. We asked community members to share their top 5 subnets, the ones they believe will define the next phase of the ecosystem. Let’s explore their picks:
Taostacker’s Picks
1. Ridges AI (SN62) – Grossly undervalued open source / decentralized player that is eating centralized competitors for lunch right now (at a fraction of the cost, too). Also paving the way and “writing the playbook” for future subnets as they think outside the box for better ways to reduce miner sell pressure and increase alpha value to investors.
2. SportsTensor (SN41) – With the Polymarket partnership, they will have ridiculous revenue coming very soon. The flywheel is strong with these guys.
3. Sundae Bar (SN121) – My AI background makes me especially interested in the AI agent subnets, and behind Ridges, I believe Sundae Bar does the agent marketplace the best way. The ‘automate your entire business’ vision and goal is ambitious enough to keep them competitive with other agent launchpads, and their unique way of handling incentives makes them stand out.
4. Zeus (SN18) – As I’ve gotten to know the team, I’ve come to appreciate that these guys love having fun while they are building out the most intelligent climate prediction tool on the planet. I expect big things out of this project, including some massive deals soon.
5. VoidAI (SN106) – The bridge to retail liquidity is VoidAI. They will have working interoperability with all major chains soon, as well as integration with Chainlink CCIP, which will be the major link needed to enable the purchase of subnets directly from CEXs. Investor buying aside, the fees and revenue from the bridge alone here will send the price flying.
Pete R’s Picks
18 Zeus – I understand it surpasses the accuracy of existing weather prediction models and has a vast and diverse market for sales. I believe one or two sales with major industry players could open the floodgates. It still has many options to expand into, in terms of geographies for miners to work on and various weather prediction parameters. Overall, it feels very much like one of the less hyped, more action-oriented subnets.
62 Ridges – What can I add that others haven’t already said? I am not a technical expert, but the statistics and quality metrics of the agents being built speak for themselves and will continue to improve. The main hurdle will be turning that into a product that competes with the big players. That’s a challenge, but from what I see of Shaq so far, I have no reason to doubt it is achievable – the decision-making seems solid. Whether that involves building a larger team or remaining small for the time being, I am confident it will be well thought out. The plans for the alpha token are also exciting, showing a real appetite for innovation and value creation, where it could be easier to rest on more traditional, already tested token plans.
75 Hippius – Mog’s recent appearance on Revenue Search was the best I’ve seen. The alphanomics are well thought out and sound great on paper. The business has several avenues for revenue, and undercutting the competition on price seems like one of the easier sells. Overall, I’m very confident in Mog and feel his connections and experience in Bittensor will be invaluable.
123 Mantis – I couldn’t not include an Atlas subnet on the list, so with 125 being largely unknown, I went with Mantis. I have no doubts that Mantis will excel in terms of the tech and quality of output, and the paths to revenue should work well. While communication could be improved, that is fixable with support. From a previous Mantis AMA, it seems clear that Atlas’s plans prioritize driving value back to Alpha.
41 Sportensor – I may be wrong, but it feels like strong revenue is just around the corner. It will be interesting to see what the Polymarket partnership looks like in practice, but my guess is that it will work out very well for the subnet. There’s no doubt that Neuro is a genius and brings incredible value in certain areas. From the outside, at least, it seems that Leo can manage that personality and is clearly delivering on the business side.
DKC’s Picks
Distributed Training (Subnet 38)
If the final preparations for the 8B model are completed and the developers’ discipline truly pays off, they‘ll be a real problem. It offers distributed training of large AI models. Uses Butterfly All-Reduce for efficient data synchronization, reduces costs, and scales training infrastructure. Instead of the miners working separately on something, they share the task and build it together.
Zeus (Subnet 18)
Decentralized climate and weather forecasting that outperforms Google in predictions. Utilizes global environmental data and innovative models. Improves forecasting accuracy efficiently and cost-effectively. Develops trustworthy environmental predictions with Bittensor integration.
Ridges AI (Subnet 62)
Automates software development via AI-driven agents on a decentralized marketplace, currently ranked 4th globally according to the SWE benchmark, surpassing centralized models, and plans an API for enterprises. Aims for massive efficiency gains. 4 Watches speak for themselves. ⌚️⌚️⌚️⌚️
MANTIS (Subnet 123)
Decentralized platform for AI-based financial market predictions with community voting for signal quality, led by the prodigy Atlas, aka Barbarian, and team. Focused on hedge funds and DeFi. Enables democratic access to high-performance forecasts and tokenized financial AI.
VoidAI (Subnet 106)
Connects Bittensor with SOL and links via a DeFi bridge. Facilitates TAO and Alpha token trading as well as liquidity provisioning. Boosts demand and ecosystem growth.
Sebbe P’s Picks
Ran some quick maths:
If Hippius (SN 75) makes $1,000,000 in profit (10% of the entire supply), which they will do very soon, they will add a lot of buying pressure. (The entire market cap rn is $11m). + People will buy to stake + miners gotta stake to compete = Veeeery high price.
And on top of that, the revenue bought alpha’s yield will benefit hAlpha holders. Then, hAlpha stakers will earn great yield + appreciation in SN 75 price.
There is simply not enough liquidity for it not to skyrocket.
I think we all can just throw away all the knowledge we thought we had about valuation out the door when this wheel starts spinning, wth.
This makes my list:
Hippius (SN 75)
Zeus (SN 18)
Vidaio (SN 85)
Cartha (SN 35)
VoidAI (SN 106)
RVCrypto’s Picks
Ridges (SN 62) – SOTA. This is going to be the first breakout subnet.
Score (SN 44) – Competes with human annotators and is creating a whole new market with their new TurboVision tech. + one of the best, if not the best, team in the Bittensor ecosystem.
Zeus (SN 18) – SOTA weather prediction. Everyone not using their weather prediction is going to pay opportunity cost, doesn’t matter in what industry, from agriculture to energy trading.
Sundae Bar (SN 121) – Perfect way of using a listed company in combination with a subnet, and incentives are spot on as well.
Quantum Subnets (SN 63/48) – before the end of the year, it is going to be clear how well connected they are and how and why they are using Bittensor. I expect a lot.
Chairman TAO’s Picks
1) Ridges is a story of David and Goliath. A group of plucky young men in their early 20’s spending less than a $1million in alpha emissions going up against a $180 billion monster that is Anthropic. It captivates the imagination of what can be done with Bittensor.
2) Score pivoted very early on from their original vision, and it’s paid off massively. This shows they are adaptable and responsive to challenges. With the goal of being the optic nerve of AI, they have a very wide scope of use cases and markets that their tech could be applied to.
3) Vidaio is not a groundbreaking subnet that will grab headline,s but it is fixated on a very specific market that is ripe for disruption. I see it as one of those so-called “boring businesses” that quietly makes everyone who’s involved in it a lot of money without reinventing the wheel.
4) Templar because they are trying to do something that goes beyond business and wealth. Their mission is to relinquish control of training LLMs from centralised entities and put it into the hands of the people. For this, I praise them very highly.
5) Hippius, because it is the subnet that can stitch so many other subnets together. Like a mycelium net that is spread throughout a forest that creates a memory of all that is happening on Bittensor.
Cryptnomad’s Picks
1) RIDGES SN 62 – Likely to be the first subnet to break out and surpass the incumbent large centralized competitor on quality, cost, time to build, and price. It will strike the first fatal blow to centralized systems and show the world what is happening on Bittensor. People will realize Bittensor is the model for how things develop: ideas, businesses, products, services. Not just AI, everything.
The following four are follow-ups among several others that show Ridges isn’t a one-off, but that there are many top-level industry-disrupting projects being built on Bittensor:
2) HIPPIUS SN 75– Cloud storage that will be an ecosystem within itself with a crack innovative team. Again, shows that products built on Bittensor outperform and outprice any existing competition by wide margins.
3) ZEUS SN 18– Beyond and above existing SOTA for weather prediction shows how the Bittensor ecosystem produces superior results in several categories. The prediction models already are outperforming all weather predictions on the planet.
4) SCORE SN 44 – Essential tech for embedded AI turning visuals into useful data. The infrastructure of the future is being built by teams like SCORE.
5) RESI SN 46 – They are building a data oracle that is essential to bring real estate into Web3, on which the future of real estate will be built. And they’re the only ones doing it because it’s the type of thing that can only reasonably be done on an incentivized, decentralized open open-sourced structure.
Be the first to comment