HashTensor’s mission is to create the most profitable crypto mining opportunities by incentivizing miners to contribute hashrate to our pools. Miners earn dual rewards and drive value back into the Alpha token through buybacks and burns.
Website: hashtensor.com
X: x.com/HashTensor
HashTensor is a decentralized Bittensor subnet designed to track and validate mining activity across multiple proof-of-work blockchains. By introducing a model called Proof of Effective Work (PoEW), the system rewards miners not just for raw hashrate, but for performance, uptime, and reliability, allowing them to earn Alpha tokens based on their actual contribution.
HashTensor captures the value of mining by redirecting hashrate into a performance-driven reward model that fuels Alpha’s growth through buybacks and token burns.
Here's how it works:
- Miners contribute hashrate to supported PoW networks (starting with Kaspa).
- Initially, all mined KAS is collected by the subnet wallet. With the next update, 95% will go directly to the miner’s wallet, and 5% to the subnet wallet as a pool fee.
- The KAS collected from pool fees is converted into TAO, then used to buy back Subnet 16 Alpha tokens from the open market.
- Each buyback burns 90% of Alpha, cutting supply and strengthening its deflationary design. This ensures the team no longer has access to those tokens.
This model transforms raw hashrate into verifiable economic output, rewarding miners while systematically increasing the scarcity and value of Alpha over time.
Current fees percentage:
- miners fee: 100%
- buy&burn ALPHA: 5%
Goal with the next update:
- miners fee: 5%
- buy&burn ALPHA: 90%
Miners are not just mining. They are producing a next-generation form of hashrate that is measurable, verifiable, and rewarded through tokenized incentives (subnet 16 Alpha).
This is the core commodity of HashTensor: validated, trust-scored hashrate.
Designed to support any proof-of-work network, the subnet is capable of integrating with multiple cryptocurrencies. The first supported coin is Kaspa (KAS). Support for additional PoW coins is planned as the subnet evolves, expanding mining opportunities across a wider range of networks.
Miners can connect their Bittensor hotkey to one or more Kaspa mining workers. The system monitors each worker’s performance and submits ratings to the Bittensor network. It is easy to configure and requires minimal setup, making HashTensor one of the easiest subnets to mine on within the Bittensor ecosystem.
- Collects real-time data from Kaspa miners using Prometheus
- Maps Bittensor hotkeys to Kaspa workers
- Validators rate miners based on their contribution
- Rewards are distributed based on ratings and uptime
- Kaspa integration as the first supported PoW coin
- Subnet launch powered by Proof of Effective Work (PoEW)
- Initial Alpha buybacks using mined KAS
- Launch HashTensor dashboard & miner leaderboard
- Start Alpha burns (initially 5% of each buyback)
- Transition to a pool-based mining model
- Progressive fee reduction from 100% to 10% miner fee
- Increase Alpha burn percentage from 5% toward 90%
- Integration of additional PoW networks beyond Kaspa
Validators compute a score for each hotkey in four steps, producing a value between 0.0 and 1.0:
- Aggregation of Effective Work
- Sum each worker’s valid shares multiplied by its difficulty:
Effective Work = ∑₍worker₎ (valid_shares × difficulty)
- Sum each worker’s valid shares multiplied by its difficulty:
- Normalization
- Identify the maximum Effective Work among all hotkeys:
max_work = max₍hotkey₎(Effective Work) - Divide each hotkey’s Effective Work by max_work to obtain a normalized score in [0.0, 1.0]:
Normalized Rating = Effective Work / max_work
- Identify the maximum Effective Work among all hotkeys:
- Uptime Penalty
- Compute the average uptime across a hotkey’s workers (each uptime as a fraction of the monitoring window, from 0.0 to 1.0).
Average Uptime = (∑₍worker₎ uptime_fraction) / (number_of_workers) - Apply a nonlinear penalty using an exponent α (default α = 2.0):
Penalized Rating = Normalized Rating × (Average Uptime)^α
- Compute the average uptime across a hotkey’s workers (each uptime as a fraction of the monitoring window, from 0.0 to 1.0).
- Clamping
- Ensure the final score remains within [0.0, 1.0]:
Final Rating = max(0.0, min(1.0, Penalized Rating))
- Ensure the final score remains within [0.0, 1.0]:
By default, a hotkey with perfect work and uptime (Normalized Rating = 1.0, Average Uptime = 1.0) will receive a score of 1.0; partial uptime or lower work reduces the rating accordingly.
Below is a high-level diagram showing how Kaspa miners, the validator, and Bittensor interact:
flowchart TD
%% ==== Miner Identity ====
subgraph "Miner"
GPU["GPU / ASIC"]
Hotkey["Bittensor Hotkey"]
end
%% ==== Kaspa Network ====
subgraph "Kaspa Network"
Kaspad["Kaspad Node"]
Bridge["Kaspa Stratum Bridge"]
Prometheus["Prometheus"]
Grafana["Grafana Dashboard"]
GPU -->|"Stratum connection"| Bridge
Bridge -->|"New block templates / submit shares"| Kaspad
Bridge -->|"Metrics (per worker)"| Prometheus
Prometheus --> Grafana
Bridge -->|"All KAS mined →"| Owner["Pool Owner Kaspa Wallet"]
end
%% ==== Bittensor Network ====
subgraph "Bittensor Network"
Validator["HashTensorValidator"]
Mapping["Hotkey <-> Worker Mapping DB"]
ValidatorAPI["Validator API / Mapping Registrar"]
Yuma["Yuma Consensus"]
SubnetBoost["Subnet Emission Boost"]
Validator -->|"Fetch metrics"| Prometheus
Validator -->|"Lookup mapping"| Mapping
Validator -->|"Send ratings"| Yuma
ValidatorAPI --> Mapping
Hotkey -->|"Register hotkey + worker + signature"| ValidatorAPI
Yuma -->|"Alpha reward to hotkey"| Hotkey
SubnetBoost -->|"Increases TAO/day"| Yuma
end
%% ==== Ownership & Emission Loop ====
Owner -->|Sends KAS| CEX["CEX / OTC Exchange"]
CEX -->|"Convert KAS → TAO"| OwnerTAO["Owner's TAO Wallet"]
OwnerTAO -->|"Stake / Buyback Alpha"| SubnetBoost
Diagram: Flow of mining, metrics, mapping, and rewards between Kaspa, Bittensor, and the validator.
❗ Before you start:
You must create a Bittensor coldkey and hotkey, then register your hotkey in the HashTensor subnet before participating.
- For the Finney network, run:
btcli subnet register --netuid 16- For the testnet, run:
btcli subnet register --netuid 368 --network testIf you skip this step, you cannot participate in the subnet.
- Modern GPU or ASIC (see mining guide for details)
- Reliable internet connection
- Sufficient system RAM and storage for mining software
Watch the full setup guide for mining on the HashTensor Subnet: [Watch the video]
| Component | Requirement |
|---|---|
| CPU | 2+ cores |
| RAM | 4–8 GB (for share analysis) |
| Network | Stable connection to pool API (low latency) |
| Storage | Minimal (just logs & lightweight data) |
Validators don't need GPUs — they are not mining.
They just analyze data and assign weights properly.
For detailed installation and configuration instructions, please refer to the guides above.
Can multiple workers be linked to one miner hotkey?
Yes. A single hotkey can support multiple workers. There’s no hard limit, miners are free to connect as many workers as they like to one hotkey.
Do validators earn KAS rewards?
Not directly. Validators receive Alpha for evaluating miner performance on Subnet 16. A future update may introduce validator revenue-sharing from the subnet’s earnings.
Is all mined KAS collected by the HashTensor wallet?
Currently, yes. With the upcoming update, 95% of mined KAS will go directly to the miner’s wallet, and only 5% will go to HashTensor as a pool fee.
When do Alpha buybacks start, and are they automated?
Buybacks will begin shortly after the subnet launch. Initially, they will be manual and publicly announced on our X page.
When and how are Alpha tokens burned?
Alpha burns will begin after the first buyback and are executed using Bittensor’s native burn extrinsic. All burns are transparent and verifiable on-chain.