@bridgemindai Who cares about closed source US labs? I'm on team OS. Glad to see a US entrant into the frontier OS with ornith 1.0 but otherwise openai and google give us table scraps with Gemma and gpt oss
@ArtificialAnlys@MiniMax_AI I dont understand why the size of the model is some big secret until they release the weights... Like you obv know what the size is gonna be. So why wait? People can't even tell what class of model without that key metric. 230b or 800b parameters makes a big diff
@melvynx Its good for its size. But yeah the chinese (and one time russian!) characters randomly thrown in does break the quality argument and a lot of use cases where that's unacceptable.