Key takeaways

• Kaito yaps paid people to post about crypto projects on X, and at the peak more than 200,000 people a month were taking part and over $95 million in rewards had been paid out.

• Kaito shut down yaps on 15 January 2026 after X banned apps that reward posting, following a day when bots produced 7.75 million crypto posts.

• A study of 40 token launches found social engagement explained almost none of a token's price performance, which is why Kaito's paid leaderboards died while its data business and curated creator program survived.

Kaito yaps turned posting about crypto into a paid job for just over a year, and in doing so put a public price on attention for the first time. Kaito, a startup that had raised $10.8 million from investors including Sequoia and Dragonfly to analyze crypto social media, announced yaps in December 2024 and turned its scoring system into the industry's attention leaderboard. At the peak, more than 200,000 people were yapping each month and over $95 million in rewards had been paid out, and attention stopped being a metaphor and became something with a ticker, a salary and a workforce.

The buyers were real and the price lists were public. dYdX published a $50,000 monthly prize pool, with $2,500 each for the top ten posters, and Arbitrum committed 400,000 ARB to its own campaign. The attention rankings even started raising money directly, with Kaito's sale platform passing $170 million in pledges, $71.5 million of it for a single sale. Then, on 15 January 2026, the whole thing ended in an afternoon.

The moment attention had an explicit price, the supply became infinite and the value went to zero.

How did Kaito yaps work?

Kaito yaps were points that Kaito's AI gave people for posting about crypto projects on X. The algorithm scored posts on quality and weighted engagement from "smart followers", accounts it judged influential, rather than raw follower counts. Projects funded leaderboards, and the top-ranked posters, called yappers, shared token or cash rewards each month.

Kaito kept the exact formula secret to stop people gaming it, and ranked both posters and projects by mindshare, which is simply a project's share of all the crypto conversation happening at a given time. That created a new category the industry called InfoFi, short for information finance, where attention and information become assets you can trade and earn from. A project could now buy a higher mindshare number by funding a leaderboard, and thousands of people had a reason to post about it every day.

Why did Kaito shut down yaps?

Kaito shut down yaps because X banned them. On 15 January 2026 X's head of product, Nikita Bier, said the platform would no longer allow apps that pay people to post, citing a surge in AI-generated spam and reply spam. Kaito's founder said permissionless paid posting was no longer viable, and the KAITO token fell about 17%.

The trigger was a flood. Pay for posts and posters show up, except more and more of them weren't people, and six days before the ban CryptoQuant's chief executive Ki Young Ju counted 7.75 million bot-written crypto posts on X in a single day, a 1,224% spike. X then cut off incentivized posting apps and Kaito's community of roughly 157,000 yappers with them, and one retrospective estimated that an InfoFi category worth roughly $649 million unwound on one platform's policy change.

The audience had already made its own call. By the end of April, Bier said crypto had become the most-muted topic on X, ahead of politics, which is the people on the receiving end grading years of paid crypto content and failing it.

Did paying for posts move token prices?

Barely, if at all. A study of 40 major token launches and more than 50,000 data points found that engagement on X explained almost none of how tokens performed, with an R-squared of 0.038, a score where 1.0 means engagement predicts price perfectly and 0 means it predicts nothing.

That research, by the Simplicity Group and reported by Odaily, tested the assumption the whole market rested on, and likes, comments and retweets even showed a slightly negative link to price. Projects were spending real money to move a number that didn't connect to anything they wanted. It's the same conclusion the crypto clipping campaigns and the rise of AI crypto influencers point to from the other direction, because attention only converts when a real person actually meant it, and a leaderboard that pays per post strips the meaning out on purpose.

What replaced Kaito yaps?

Kaito replaced yaps with Kaito Studio, a selective marketplace that matches brands with vetted creators on X and other platforms and pays for agreed work rather than automated points. It opened in beta in February 2026 with 16 brand partners, while Kaito's paid data product for professional teams kept running.

Kaito's data business, the part that measures attention instead of paying for it, was reportedly making about $33 million a year from institutional subscriptions, and the company relaunched with a curated program of 16 chosen brand partners instead of an open leaderboard. That sequence is the clearest lesson in the story. Measuring attention turned out to be valuable and durable, while paying for attention by the post, in public and at unlimited scale, produced exactly the kind of engagement nobody could use.

What we'd tell a founder about Kaito yaps and paid attention

Use attention data as a dashboard, not a campaign. Knowing who is gaining mindshare, which stories are rising and where your project sits is useful research, so watch it. Just don't confuse moving your own number with gaining users, because the research says those are two different things.

Tie any paid posting to what happens after the post. Activity should qualify someone and results should get them paid, through tracked links, sign-ups, deposits and users who stay. A program that pays purely for visibility is an open invitation to the 7.75-million-post bot machine.

Never buy volume in a feed your buyers can mute. Every low-effort paid post makes the whole topic easier to ignore, including your own posts. The bar isn't whether a post got views, it's whether someone would be glad they follow the person who wrote it.

In creator programs, curation beats scale. The version that survived is 16 vetted partners, not 157,000 anonymous farmers. A small roster you actually choose, paid on results and free to tell you no, is the difference between borrowed trust and rented noise.

It's tempting to file the yap era under crypto absurdity, but no other industry has produced this data about what happens when attention becomes directly tradable. The supply explodes, machines outcompete humans, the link to business results measures at almost zero, and eventually the platform pulls the plug. Attention matters as much as it ever did, and the experiment simply proved you can't buy it by the pound.

Frequently asked questions

How does Kaito yaps work?

Kaito yaps no longer work, because the program ended on 15 January 2026. While it ran, Kaito's AI scored posts about crypto projects on X, weighted engagement from influential accounts, and ranked posters on project-funded leaderboards. Top yappers shared monthly rewards, such as dYdX's $50,000 pool that paid $2,500 each to its top ten.

Can you still earn Kaito yaps?

No. Kaito sunset yaps and its incentivized leaderboards after X banned apps that reward posting in January 2026. Its successor, Kaito Studio, is a selective marketplace where brands hire vetted creators for defined work, and it launched in beta with 16 brand partners rather than open, points-based rewards.

What is InfoFi in crypto?

InfoFi, short for information finance, is the idea of turning attention and information into assets that people can trade or earn from. Kaito's yaps were the best-known example, paying more than $95 million to posters by mid-2025, and the category was valued at roughly $649 million before X's January 2026 ban.

What is mindshare in crypto?

Mindshare is a project's share of all the crypto conversation happening at a given time, usually measured across posts on X. Kaito popularized it as a leaderboard metric. A study of 40 token launches found engagement explained almost none of price performance, so mindshare is useful research but a weak target on its own.

Why did X ban InfoFi apps?

X's head of product, Nikita Bier, said the platform would stop allowing apps that pay people to post because they drove a surge of AI-generated spam and reply spam. Six days earlier, bots had produced about 7.75 million crypto posts on X in a single day, a 1,224% jump.