Best LLM for 8 GB VRAM · 2026

Best LLM for 8 GB VRAM: what fits, and what to run

A 8 GB card hands a runtime about 7.0 GB once the display and driver have taken their share. At Q4_K_M with an 8K context, 28 of 80 models in the catalogue fit — 22 with headroom, 6 tightly. The pick for most people is Gemma 4 E4B: Laptop class with images and audio in. Speeds are estimated for a GeForce RTX 4060; the fit verdicts are the same for every card in the class.

Which one should I run?

One pick per job, from the editorial shortlist, checked to fit at Q4_K_M and 8K. How picks are chosen.

The long-context row is judged at 128K: Qwen3.5 2B holds that window on 8 GB at Q4_K_M with 1.50 GB of KV cache — that is why it can differ from the best overall pick.

Everything that fits 8 GB

Largest first. "Max context" is the longest window the card holds at that quantisation with an f16 cache; a q8_0 cache roughly doubles it.

Just out of reach

These run with part of the weights in system RAM (32 GB assumed), at a few tokens per second. Each needs a bigger card to run properly — the link says which.

Every total is quantised weights + f16 KV cache at 8K + 0.6 GB runtime overhead, against 7.0 GB usable. Tokens per second are estimated from memory bandwidth and never measured. The method.