ANTS

알파를 듀오 공지방1d ago

AntSeed Research: How to Use AI 90% Cheaper (Why DCG, Ventrice, Spark Invested?) I really hope everyone who worries about AI limits sees this once. There's a real way to use AI reasonably while farming promising $ANTS token! AntSeed is a project connecting AI token suppliers <> demanders, supplying tokens at crazy cheap discount rates of 90-98% for latest models like Opus 5.5 and Astra. Spark Capital, which led Anthropic Series C at $1.5B valuation (currently 366x lol), plus DCG and Venice—top-tier VCs with real credentials invested $2.4M into the $ANTS token round, and I genuinely think there's an angle for me here. Honestly just using discounted tokens well is very efficient already, but with discount token farming early on looking good on compound efficiency, I've decided to just spend the next month farming AntSeed. - Exploding AI infrastructure market - AntSeed's core investor lineup - How it works - Tokenomics analysis - Optimal farming strategy. https://blog.naver.com/money0stack9/224457439082

183Spread to 2 channelsShow Korean original

무사시의 코인베기2d ago

I did some airdrop farming for Antseed back in May, and after calculating, the mining cost then and now doesn't seem that different? Actually, USDC volume seemed bigger back then. My take: as inference markets keep growing and VC-backed projects enter crypto, in terms of founder career prospects, this one seems pretty solid. Among on-chain projects there's Surplus, and their token consumption volume is really solid, but since they haven't turned on the fee switch yet, we'll need to see the valuation after they do. $ANTS that's been farmed is only tradable in staked form (lANTS), and while volume isn't huge, it's trading around 100M. Looking at the actual token consumption in the platform and committed USDC, I honestly wonder if this valuation makes sense. Korea's been getting ad pushes + viral moments that feel like the weight class went up too much? The mining cost is roughly 26M if using a model that gets full leverage rewards. The current mining cost seems reasonable compared to other inference projects. Farming is either by supplying or using models, and I think just attaching a model to the agent you're already using is the most convenient. Rather than dumping it all at once, it looks better to gradually unwind. You can also stake (lock up) farmed tokens and earn yield on top. If you're thinking about farming, it'd be good to read the above post first https://antseedstats.com/ https://antseedpoints.com/

592#AIAgentsShow Korean original

정열맨김봉봉2d ago

I'm thinking of farming this one because I think it's pretty decent. (Not an ad) 1. Antseed is OpenRouter model of AI model suppliers <> mediating between users, which is a decentralized + crypto version. If you use the AI models you normally use through this, you can mine ANTS. 2. OpenRouter is a company that lets you pick and use various AI models like GPT, Claude, Gemini through one API. Users charge up credits and can use each different model. 5.5% of charged amount is the fee, and annualized revenue as of August is estimated at ~$160M. 3. Stripe acquired OpenRouter for $7B in August (OpenRouter's company valuation was $1.3B in May). Since Stripe is a payment company, there's a diverse corporate pool, and attaching OpenRouter to it lets companies use AI at lower costs with more suitable models, cutting expenses, which is why Stripe acquired them. 4. Successful business models like OpenRouter are already verified in the market, their scale is growing, and they're proving necessity through model usage amounts. OpenRouter-passed model usage went from $10M annualized in Oct 24 to over $100M by May 25, over 10x in 7 months, now processing over 10T tokens daily. 5. The idea is to bring this successful business model to crypto and decentralized version and implement it, which is what Antseed is. 6. Besides that, personally I think Antseed's composition itself is quite solid. CEO Amos worked as dealing head at eToro (original web2 FOMO, social trading app, the thing where pressing $ticker on X shows price), and Amos also participated in Colored Coins (early token concept putting different assets on Bitcoin, Vitalik also authored it) first proposed by eToro's CEO Yoni in 2012. 7. After that Amos co-founded blockchain company Colu in 2014 and was CEO, where Spark Capital which led eToro's early rounds also led Colu investment, and DCG (then predecessor BOC) also invested together. 8. In this case, Spark Capital led Anthropic's ~$450M Series C which we know well, and the partner also joined Anthropic's board. Early investor in Twitter as well. Previously invested companies like eToro

1KSpread to 2 channelsBitcoin#ListingShow Korean original

Manbull7d ago

This is an article on Onchain Inference Market compiled by Bankless, and I think it's worth a read for those of you holding on-chain AI exposure. The content itself is brief and reads more like an explanation of customers/targets/structure/principles of on-chain projects. The projects mentioned seem to be Orbio, Surplus, and Antseed. They don't just explain the advantages of each, but also the weaknesses, which makes it good for objective understanding. In rough summary: Orbio: Does partnerships well and ecosystem building well, but the challenge is whether transaction fee-based credits can fundamentally be sustained. Surplus: A platform selling unused subscription quantity, highest token usage + popularity. However, it has a large token supply, no revenue, and is mainly used for cheap AI models. (The connection to the token was weak, but is expected to improve after migration.) Antseed: Similar to Surplus but uses its own gateway while preventing fraud with AI models (claiming DeepSeek v4 but turns out it's v3). However, a weakness is inflation since ANTS is distributed as rewards in the tokenomics. https://x.com/Bankless/status/2115049497373921667 For now, these 3 are definitely the most mentioned and receive the most attention in AI Inference.

1.5KSpread to 2 channelsShow Korean original