Deep Dive
1. Reward System Overhaul (October 2025)
Overview: Grass updated its incentive model by splitting Grass Points into Uptime Points and Network Points. This change makes rewards more closely tied to how much a user's bandwidth is actually used by the network, rather than just connection time.
The new system distributes 1 million Network Points daily based on bandwidth utilization, connection stability, and regional demand. A unified Rewards Dashboard now lets users track both point types along with referrals. This update coincided with Epoch 12, which ran from 30 September to 31 October 2025.
What this means: This is bullish for GRASS because it creates a fairer and more transparent earning model. Users who provide more valuable, stable bandwidth can earn more, which should improve network quality and attract serious participants. It directly ties token rewards to real-world network utility.
(Source)
2. Sion Technical Upgrade (February 2025)
Overview: The Sion upgrade was a major backend improvement that significantly boosted the network's data processing capacity and scalability to handle the demands of AI companies.
It enabled the network to process multimodal data—including text, images, and 4K video—at a much larger scale. The upgrade introduced horizontal compute scaling, raising the network's data handling capacity to over 1 petabyte per day. A key milestone was scraping over 1 million GB of public web data in a single day shortly after the upgrade.
What this means: This is bullish for GRASS because it strengthened the project's core value proposition: providing reliable, large-scale web data for AI training. By improving technical infrastructure, Grass becomes more competitive against centralized web crawlers, supporting long-term demand for its services and the GRASS token.
(Source)
Conclusion
Grass's development trajectory shows a clear focus on enhancing both technical infrastructure and user incentive alignment, strengthening its position as a decentralized AI data layer. Will continued refinement of its reward mechanics further accelerate network growth and data quality?