And the fourth one I want to mention is solana:BXoHJddsWJLHtAopeiSbKUSELsu8hSFMs8baGMDkpump
This is one of the newer AI coins on Solana that caught my attention recently. Not just because of the price action, but because there is already something real running behind the token
That said, with solana:BXoHJddsWJLHtAopeiSbKUSELsu8hSFMs8baGMDkpump, I think it is important to separate what has actually been shipped from what is still just a thesis
1. What is CrawlNet actually building?
solana:BXoHJddsWJLHtAopeiSbKUSELsu8hSFMs8baGMDkpump is the token behind CrawlNet, an experiment attempting to build a crypto-native LLM using data collected directly from Web3
Instead of simply launching a token and attaching an AI narrative to it, CrawlNet is operating a network of crawlers. These crawlers read whitepapers, documentation, governance pages, forums, blogs, and other crypto-related sources
The data is then cleaned, deduplicated, and added to a dataset used to train the project's model, Queen
What I find interesting is that most of this process can actually be tracked through the dashboard. In recent snapshots, there were more than 80 crawlers operating, with over 12,000 pages crawled across hundreds of crypto-related domains
2. Queen is the most important part of the thesis
Queen v1 was trained from scratch using a nanochat-style approach, with roughly 42.1M parameters and a corpus of around 26.2M tokens collected from 10,521 pages
The team has also published the Queen v1 weights on Hugging Face and released an experimental chat interface
Meanwhile, data collection is continuing for Queen v2. The dataset recently reached around 32M tokens, with the team targeting roughly 60M
This is what makes solana:BXoHJddsWJLHtAopeiSbKUSELsu8hSFMs8baGMDkpump more interesting to me than many other AI memes
The team is not simply saying they are going to build AI at some point in the future. The crawlers are running, data is actually being collected, and the first model has already been trained
But it is also important to keep the scale in perspective
42M parameters is still a very small model, and the dev himself has said that Queen is still in testing. There is currently no independent benchmark showing that Queen can outperform existing small models
So for now, I view Queen v1 more as a proof of concept than a finished AI product
3. How does solana:BXoHJddsWJLHtAopeiSbKUSELsu8hSFMs8baGMDkpump fit into the system?
This is probably the part of the project design I find most interesting
To spawn a crawler, users need to burn solana:BXoHJddsWJLHtAopeiSbKUSELsu8hSFMs8baGMDkpump. The initial cost was around 200,000 tokens per crawler, although the team later adjusted the economics and reduced the spawning cost
Once created, crawlers continue collecting crypto data and contributing it to Queen's dataset
Creator fees generated from trading flow into the treasury. Under the current structure, roughly 60% goes toward compute and 40% goes into the reward pool for crawler owners
This split was adjusted during the first few days, as earlier updates had allocated a smaller percentage to crawler owners
The reward pool settles every 12 hours. To qualify for rewards in a round, a crawler needs to bring in at least 25 new pages during that round
In one of the early rounds, the team reported distributing roughly 65.2 SOL across 69 crawlers owned by 36 wallets. At that point, total creator fees generated since launch were around 142.9 SOL
There are also refuel and order mechanisms
Once a crawler reaches its earn cap, more solana:BXoHJddsWJLHtAopeiSbKUSELsu8hSFMs8baGMDkpump can be burned to refuel it
With orders, users can burn around 10,000 solana:BXoHJddsWJLHtAopeiSbKUSELsu8hSFMs8baGMDkpump to request that a crawler reads a specific website, with a limit of roughly 50 pages from a single host
However, these orders need to be approved by an operator before execution, so this part of the system is not yet fully automated on-chain
4. The flywheel is interesting, but still heavily dependent on attention
In simple terms, CrawlNet currently works like this :
Trading generates creator fees
Fees fund compute and crawler rewards
Users burn solana:BXoHJddsWJLHtAopeiSbKUSELsu8hSFMs8baGMDkpump to spawn or refuel crawlers
Crawlers collect more data
The data is used to train Queen
Part of the token supply continues to be burned
During the early stage, the team reported that roughly 2.7% of the supply had been burned. However, this was only a snapshot at that point in time, and crawler economics have since been adjusted, so I would not treat 2.7% as a fixed figure for the current state of the system
The bigger issue is that crawler rewards are not currently funded by revenue generated by Queen itself
Most of the money still originates from solana:BXoHJddsWJLHtAopeiSbKUSELsu8hSFMs8baGMDkpump creator fees
High volume means a larger reward pool
Lower volume means a smaller pool
Smaller rewards mean less incentive to spawn and operate crawlers
So in its current form, I still see this as an attention flywheel attached to an AI experiment rather than a complete business flywheel
For this thesis to become significantly stronger, I would want to see actual users of Queen, third parties finding value in the dataset, or revenue streams that do not depend entirely on trading volume
5. The dev is another part I will be watching
The developer behind CrawlNet is Matthew,
@matthewabides, who also operates
@crawlnetwork
Before CrawlNet, Matthew worked on Swarm, an AI agent project on Robinhood Chain. Swarm ran for more than two weeks before he decided to shut it down and withdraw the dev fees, after which he moved on to CrawlNet
Because of that, I would not use the dev's track record as a reason by itself to trust solana:BXoHJddsWJLHtAopeiSbKUSELsu8hSFMs8baGMDkpump
For this project, I care much more about what Matthew continues to ship after the initial hype dies down
Another thing worth clarifying is that the crawler visualization itself is not an entirely original concept. Matthew has acknowledged inspiration from another crawler visualization, and another token with a similar narrative appeared before solana:BXoHJddsWJLHtAopeiSbKUSELsu8hSFMs8baGMDkpump
So I would not value solana:BXoHJddsWJLHtAopeiSbKUSELsu8hSFMs8baGMDkpump based on a "first mover" thesis
What matters more is that the team has turned the concept into an actual working system with crawlers, token burns, rewards, a growing dataset, and a model
6. Price action
solana:BXoHJddsWJLHtAopeiSbKUSELsu8hSFMs8baGMDkpump launched at a very low market cap and quickly became one of the more talked-about coins within this narrative
ATH reached roughly $7M to $8M market cap, while the chart is currently sitting around the $5M area after a very aggressive expansion
During the early phase, 24h volume climbed above $15M and the holder count reached around 7,000+
That is a significant amount of attention for such a new token
However, liquidity is only in the several-hundred-thousand-dollar range. Relative to its market cap and trading volume, liquidity remains fairly thin, meaning relatively modest buys or sells can move the market cap aggressively in either direction
Because of that, I would not chase expansion candles here
7. What are the next catalysts?
For me, the most important catalyst for solana:BXoHJddsWJLHtAopeiSbKUSELsu8hSFMs8baGMDkpump is not another KOL call
It is whether the team keeps shipping
Queen v2 is currently continuing to collect data, with a target of around 60M tokens
The team has also said that weights from future training runs will continue to be published on Hugging Face. There are also planned updates to crawler economics and the harness to make interacting with Queen easier
If v2 ships, the dataset continues growing, and most importantly, people outside the team begin actually using Queen, the thesis becomes significantly stronger
On the other hand, if the crawlers continue running but Queen fails to generate meaningful utility, the market may eventually value solana:BXoHJddsWJLHtAopeiSbKUSELsu8hSFMs8baGMDkpump as an AI meme with a better-than-average demo rather than an actual AI infrastructure project
8. My view
The reason I am paying attention to solana:BXoHJddsWJLHtAopeiSbKUSELsu8hSFMs8baGMDkpump is pretty simple
The team has already shipped more than most new AI memes I have seen recently
The crawlers are running
Tokens are actually being burned to participate in the system
Rewards have already been distributed in SOL
The dataset is growing
Queen v1 has been trained and its weights are public
But I also do not want to overhype it
Queen is still very small, there is no independent evaluation yet, the economics remain heavily dependent on trading volume, and the dev does not have a strong enough track record for me to make a long-term bet based purely on promises
For now, I see solana:BXoHJddsWJLHtAopeiSbKUSELsu8hSFMs8baGMDkpump as an experimental AI coin with a real product worth watching, rather than a proven AI infrastructure play
After running from a very low market cap to around $7M to $8M and now sitting around $5M, I would rather watch for a pullback than FOMO into it
If Queen v2 actually ships, the crawler network continues expanding, and most importantly, real usage starts appearing outside of speculation, then the solana:BXoHJddsWJLHtAopeiSbKUSELsu8hSFMs8baGMDkpump thesis becomes much more interesting
BXoHJddsWJLHtAopeiSbKUSELsu8hSFMs8baGMDkpump