Closed Alpha Testing Phase #3

The Cortensor Closed Alpha Testing Contest (Phase #3) is a major step in expanding the decentralized AI network. This phase brings new enhancements, increased participant capacity, and a refined reward system. Early supporters and node operators continue to play a critical role in building a more scalable and robust AI infrastructure—while earning $COR rewards for their contributions.

Contest Rules

Stake Requirement

  • New Participants: Stake 4,000 $COR per node on Pool 3 to join Phase #3.

  • Pre-approved Testers: Participants from Phase #1 & #2 are automatically eligible and do not need to re-stake.

Node Uptime

Keep your nodes online and responsive for 30–45 days to qualify for full rewards.

Bug Reporting

Active bug reporting and feedback are required. Participants must:

  • Communicate with Cortensor admins

  • Report bugs or performance issues

  • Cooperate in troubleshooting

Failure to report or troubleshoot may lead to disqualification.

Disqualification Rights

Cortensor reserves the right to disqualify any participant at its discretion for:

  • Misreporting node behavior

  • Abusing the ranking or staking system

  • Acting against the community’s best interests

Long-Term Commitment

Running stable nodes over time increases your chances of being eligible for:

  • Future airdrops

  • Testnet & Mainnet whitelisting

Rewards

Rank

Total Reward

🥇 1st

$50

🥈 2nd

$50

🥉 3rd

$50

🏅 4th

$50

🎖️ 5th

$50

6th–10th

$35 each

11th-20th

$25 each

21th-30th

$15 each

31th-100th

$10 each

101th-150th

$5 each

Base Rewards

  • $20/month per active node with valid uptime and participation.

Rank-Based Bonus Rewards

  • Ranks 1–100: Bonus rewards based on performance.

  • Ranks 101–150: Flat $5 reward per qualifying participant.

Multi-Node Bonus Boost

  • 3+ nodes → 10% bonus

  • 5+ nodes → 20% bonus

Note: This multi-node bonus is applied to new nodes once during the contest period and not on a monthly basis. This is subject to change.

Role Bonus

Participants with verified roles (e.g., miner, oracle, verifier) earn an additional 5–10% on top of total rewards. Note: The total bonus from multi-node and role-based boosts is capped at 40–50% to ensure fair distribution and sustainability. This is subject to change.

Reward Tier Limits (Fair Distribution)

To maintain fairness, participants are limited in how many rank reward spots they can hold:

Rank Range

Max Spots per Operator

1st – 5th

1

6th – 10th

1

11th – 20th

1

21st – 30th

1

31st – 100th

10

101st – 150th

Unlimited

Example Calculations

Scenario 1: Running 3 Nodes (With Role Bonus)

  • Base reward: $20 × 3 = $60

  • Multi-node bonus (10%): $6

  • Subtotal: $66

  • Role bonus (10%): $6.60

  • Total: $72.60 + additional rank rewards

Scenario 2: Running 5 Nodes (With Role Bonus)

  • Base reward: $20 × 5 = $100

  • Multi-node bonus (20%): $20

  • Subtotal: $120

  • Role bonus (10%): $12

  • Total: $132 + additional rank rewards

Contest Mechanics

Max Nodes Per User

Each participant may operate up to ~25 nodes.

Winner Determination

Participants will be ranked using multiple performance metrics:

  • NodeStats Performance: Uptime, responsiveness, and throughput

  • Ping & Global Ping: Regional and global latency

  • State Machine Metrics: Quality in Create, Prepare, Precommit, Commit states

  • Task Counters: Number and accuracy of completed inference tasks

  • Bug Reports: Additional points for helpful reports

Testing Focus

This phase continues refining Cortensor’s decentralized AI infrastructure, emphasizing improvements in reliability, user experience, and scalability. It also builds upon key architectural elements introduced in previous phases.

Core Testing Areas:

  • Node Reputation Tuning & Iterations – Ongoing optimization of reputation scoring models for fair and effective miner selection.

  • Node Pool Logic (Ephemeral Sessions) – Enhancing ephemeral node reservation, assignment, and release cycles critical for routing user requests.

  • Session UI & Integrated Task Flow – Testing seamless interactions between Session, Session Queue, and miners through user-facing workflows and updated interfaces.

  • Desktop App Stress-Testing – Validating auto-install and launch performance across major operating systems (Windows, Linux, macOS) to support wide-scale participation.

Key Outcomes:

  • Improve accuracy and fairness in miner/node reputation scoring.

  • Validate ephemeral session logic and reinforce the reliability of node rotation.

  • Enhance user experience and end-to-end task handling across UI and backend flows.

  • Ensure installation and runtime stability across different system environments for broader network accessibility.

Join the Contest

Let’s build the future of decentralized AI together! 🌐

PreviousClosed Alpha Testing Phase #2NextClosed Alpha Testing Phase #4

Last updated