e1n00r / tinyserveView on GitHub
30 tok/s for 20B MoE on 8 GB VRAM. Flat throughput to 32K context. Native MXFP4 + GGUF Q4_K/Q5_K/Q6_K via ggml CUDA kernels — zero dequant. Expert offloading for models that don't fit in GPU memory.
19Apr 7, 2026Updated 3 months ago

Alternatives and similar repositories for tinyserve

Users that are interested in tinyserve are comparing it to the libraries listed below. We may earn a commission when you buy through links labeled 'Ad' on this page.

Sorting:

Are these results useful?