This work distills many-step diffusion models into few-step samplers by matching conditional expectations along the denoising trajectory. Multi-step distilled models improve over one-step variants and can outperform their original teachers on image generation benchmarks.