stratobomber
655 posts








I've built an ecosystem of AI utility around solana:EN2nnxrg8uUi6x2sJkzNPd2eT6rB9rdSoQNNaENA4RZA Level5.cloud allows you to use the token to buy perpetual AI inference credits. UsePod.ai let's you resell those credits, as well as your DIEM and MOR inference, on a secondary inference market. Inference users get all the same frontier and open-source models at a 40% discount.


It is my pleasure to introduce UsePod.ai, my entry into the @colosseum hackathon. UsePod is a two-sided marketplace for LLM inference. Independent operators run open-weight models on their own GPUs or bring their keys to other services (OpenRouter, Venice, etc.), set their own per-token prices, and earn USDC on every request served. Users send standard OpenAI- or Anthropic-compatible requests; they use the same SDKs, same base URL pattern. Centralized providers stay on as an always-available fallback. Settlement is per-request, on @solana, no token required.


leopold aschenbrenner turned $250M into $5.5 in one calendar year. his 13F just dropped and his going all in on COMPUTE and facilitators of it. > so am i < base has $VVV $DIEM $POD solana has solana:EN2nnxrg8uUi6x2sJkzNPd2eT6rB9rdSoQNNaENA4RZA and it is a hyperscaler / the only compute marketplace on solana.

@goomba_stack @SwiftOnSecurity I strongly suspect you’ll see AI inference compute markets—potentially even futures—trading at one later below the model in the inference stack. Think a liquid contract market so you could buy, sell, and trade 100 H100s in a US datacenter for Q4.

BYOK suppliers on UsePod.ai now get more pricing visibility. You can see the competitive market rate for each model, compare it against your upstream cost, and adjust your ask accordingly.








It is my pleasure to introduce UsePod.ai, my entry into the @colosseum hackathon. UsePod is a two-sided marketplace for LLM inference. Independent operators run open-weight models on their own GPUs or bring their keys to other services (OpenRouter, Venice, etc.), set their own per-token prices, and earn USDC on every request served. Users send standard OpenAI- or Anthropic-compatible requests; they use the same SDKs, same base URL pattern. Centralized providers stay on as an always-available fallback. Settlement is per-request, on @solana, no token required.


@zrpintern @dharmjack01 @squire_bot got it, squire_bot angle on monetizing idle compute makes sense. that marketplace play for unused inference credits is where the infrastructure gets interesting


