Mistral Small 4 119B 2603

Mistral Small 4 119B 2603 is a 119.4-billion-parameter sparse Mixture-of-Experts model from Mistral AI, activating 6.5 billion parameters per token across 4 of 128 routed experts plus 1 shared expert. It supports vision input, tool calling, code generation, and toggleable reasoning effort across 24 languages. A 256K context window, 40% lower latency, and 3x throughput compared to Mistral Small 3 make it well suited for high-volume agentic and multimodal workflows. Its MoE sparsity quantizes efficiently to GGUF for self-hosted deployment on multi-GPU setups.

Mixture-of-Experts

This is a Mixture-of-Experts model. MoE models load all parameters but activate only a fraction per token — learn why this changes VRAM planning, quantization, and deployment strategy.

Read the MoE deployment guide →

Hardware Configuration

Vendor Select vendor...

Product —

Platform Select platform...

Family —

Model —

VRAM —

System RAM (GB) Optional — for precise deployment recommendations

Clear Configuration

Quantization	Quality	Size	Fit
Q8_0	High	117.78 GB	—
Q6_K	High	90.96 GB	—
Q4_K_M	Medium	67.21 GB	—

Last updated: March 17, 2026