An analysis of two diffusion language models found that generation cost varies substantially across tasks and languages because the models decide during generation how many forward passes to spend, so two outputs of the same token length can differ widely in cost. The study examines the underlying factors behind that cost variation, noting that diffusion language models can write several tokens in one forward pass.
Diffusion language models can write several tokens in one forward pass, and they decide during generation how many passes to spend. Two outputs of the same token length can therefore vary a lot in cost. To understand the underlying factors of this cost variation, we analyze two diffusion models acro
EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.