# vLLM adds day-0 support for Mistral Large 4 with DSpark and NVFP4

> Original title: Day 0 support for Mistral Large 4 on vLLM

- Source: vLLM Blog
- Published: 2026-10-29T00:00:00.000Z
- Bad Influence: https://badinfluence.news/items/np42iyocgets895a4xd8p9j4c
- Original: https://vllm.ai/blog/2026-10-29-mistral-large-4

## Summary

vLLM announces day-0 support for Mistral Large 4, a 1-trillion-parameter mixture-of-experts multimodal model with 49 billion active parameters per token. The post reports about 2x output throughput on TP8 with the built-in DSpark drafter and about 1.5x more KV cache with the NVFP4 checkpoint, which stays within 2.1 points of FP8 on evaluated tasks.

## Why it matters

The post gives specific architecture, memory, and throughput figures for running Mistral Large 4 on vLLM, useful for judging serving tradeoffs and deployment layouts.