Grass scraped over 57,000,000 GB of public web data in Q1. Since the Sion upgrade, scraping volume has grown 320x.
In March alone, the network averaged 1,100,000 GB per day. That’s a scale that once belonged only to the largest companies on the internet.
Update on Android:
Initial testing with a small group of alpha users is nearing completion.
Final backend refinements are being worked on to improve reliability before the slow rollout starts.
The infrastructure for working with video data at scale barely exists. Grass Video Search is our first step in changing that.
You can now find anything in a video based on the contents of its frames, not just transcripts or tags.
Over the last several months, the Grass Foundation has quietly assembled one of the largest video datasets in history: over a billion videos from across the public web, and is segmenting them by scene with VLM annotations of keyframes. This dataset is now searchable, so you can find the exact moment a visual, pattern, or concept appears.
Curation has always been the bottleneck in video analysis. We built a system that automates it.
This is just the beginning. Over 4 billion additional videos are queued for processing, and the index is growing daily.
We’re opening access to a limited set of collaborators. If you’re a researcher, journalist, or nonprofit working with video, archives, or large-scale analysis, we’d love to hear from you.
New network ATH: more than 1,000,000 GB scraped in a day.
In the last 24 hours, Grass has gathered more multi-modal data than a human brain will process in four decades.
The claim period for GRASS Airdrop One has been extended. The extended claiming window will end on block 329341917 or roughly March 27, 2025. To be eligible for Airdrop One, you must have earned 500+ Grass Points in any epoch and linked your wallet on the Grass dashboard before October 14, 2024, at 20:00 UTC.
Claim: https://t.co/otBzRhZLLd
2025 is shaping up to be a defining year for Grass. Before diving into what’s ahead, we want to formally wrap up 2024.
In 2024, Grass:
• 0.2M users to 3M (15x)
• 8,000,000GB+ of total web data scraped
• Ability to scrape 100TB+ of web data daily
• 3 billion videos (1000x)
• Over $196 million* distributed to over 2.2 million users in one of the most widely distributed airdrops in history.
*Based on the average price of GRASS over the month of December.
Last week we had our first founder seminar of 2025. Here's what @0xdrej outlined for Grass' future:
• Sion Phase 1 results: Improving network efficiency with 60x faster data retrieval. https://t.co/OqjTIha8hi
• Multimodal search: Enable AI models to semantically search insights from video, audio, and text data in real time.
• New distribution methods: Expanding access through mobile and other platforms, making it easier for users to join and contribute to Grass.
• Deprecating the Chrome Grass extension: Transitioning fully to the more powerful and efficient desktop node
• Introducing a new product: Horizontally scaling Grass… 🤫
If you missed the latest seminar, don’t worry, future sessions will be brought to X. Below are some great breakdowns from the community:
The community nailed it with this breakdown of Grass’s multimodal search initiative. https://t.co/6jFmvPTxEP
Sergeant Lazo bringing the seminar summaries as usual! https://t.co/zzAFCG33pq
Our growing Reddit community joined in by breaking down the key roadmap points. https://t.co/xPocXRRllW
And our very own Dan pulling some of the key sections Andrej covered. https://t.co/Qx4cvENb9r
Grass is just getting started. After a year of scaling and breakthroughs, 2025 will push the limits of what’s possible, solving massive problems in AI and driving real-world impact.
None of this would be possible without the incredible support and contributions from our community. Together, we’re building a more open internet.
Grass, Ontocord, and LAION launch VALID
We’re excited to share the release of VALID (Video-Audio Large Interleaved Dataset) by the world-renowned teams at @ontocord and @laion_ai. VALID was built using the Grass Video Repository.
This dataset is comprised of 30 million audio snippets interleaved with images and text, making it the FIRST EVER video-audio interleaved dataset.
Stage 2 of the network is here.
Together, we’ve set the stage to reclaim the public web. Imagine a world where every web search is routed through the Grass Network.
The focus is now on capturing the web by building a user-owned map of the internet. The next season introduces more rewards and new product releases, such as Live Context Retrieval (LCR), hardware, the rollout of the Grass mobile app, and _______.
As Stage 2 kicks off, we’re also reflecting on the success of Stage 1. The Grass community’s impact has been incredible: 2.8 million users across 190 countries, the most widely distributed airdrop on Solana, and 100TB of web data scraped daily.
With so many AI models out there, which ones will last? Check out this 17-second clip to see how Grass’s Live Context Retrieval (LCR) engine can give models a unique edge through access to real-time, unbiased web data.
Catch @0xdrej’s discussion with @alpha_pls on Grass and Live Context Retrieval (LCR). Learn how LCR gives AI access to real-time web data, enhancing AI training and inference.
Full thread below 👇
Join @0xdrej and @Edward__Park today at 10pm (KST) as we dive into what's next for 잔디. Plus, we're giving away @GigaBuds_ to three people who like, share, and join the space!
Don't miss out, tune in here: https://t.co/HSaHkpLkfo
When AI gives bad advice, it can have real consequences.
Grass is improving the way AI interacts with information by providing tools that enable access to accurate, verifiable web data. Join millions of users and download Grass today.
We’ve just sent out an official email from [email protected] regarding the airdrop checker and wallet connect deadline. All future email updates will come from the https://t.co/fZ5YAEgdrv domain. If you’re ever unsure, ask in the Grass community Discord.