data centers are a broken model.
by the end of 2025, 5 companies owned about 71% of the world's ai compute, up from 63% in early 2024 (source: epoch ai).
everyone else rents from them, at their price and on their terms.
@LightchainAI runs ai on people's own gpus instead.
each operator brings their own gpu and stakes 50,000 LCAI. the chain picks who serves each job, and every answer leaves a public record on the job id, the transaction and the worker.
the smallest model on the network runs on a gpu with 8 gb of vram.
they also just opened its developer api on mainnet.
it works with openai's chat api, so an app can switch to open models like gpt-oss, qwen or llama by changing three things:
> the base url
> the api key
> the model name
you sign in with a wallet, not a sign-up form. it's text only, no images or tools.
lightchain's release on decrypt:
workers serving right now:
try the chat: