Comment by paxy 12 hours ago Because they have been making inference a lot more hardware efficient. 2 comments paxy Reply julianlam 9 hours ago By using Chinese research InsideOutSanta 7 hours ago I can't believe the actual answer is so far down the thread. All praise to our overlord DeepSeek for ironically helping American labs make cached tokens much cheaper.
julianlam 9 hours ago By using Chinese research InsideOutSanta 7 hours ago I can't believe the actual answer is so far down the thread. All praise to our overlord DeepSeek for ironically helping American labs make cached tokens much cheaper.
InsideOutSanta 7 hours ago I can't believe the actual answer is so far down the thread. All praise to our overlord DeepSeek for ironically helping American labs make cached tokens much cheaper.
By using Chinese research
I can't believe the actual answer is so far down the thread. All praise to our overlord DeepSeek for ironically helping American labs make cached tokens much cheaper.