localLLMriglocal, on your table

estimate

Can I run Llama 3.1 8B on an RTX 5090?

Yes, fully in VRAM (estimate)

Can I run Llama 3.1 8B on an RTX 5090? Yes, fully in VRAM (estimate). About 6.7 GB at Q4_K_M (estimate).

Updated . The later of the hardware price date and the formula date.

Llama 3.1 8B on RTX 5090, formula estimate
QuantContextWeightsKVOverheadTotalUsableFitEst. tok/s
Q4 Q4_K_M8,192 tokens4.8 GB estimate1.1 GB estimate0.8 GB estimate6.7 GB estimate31.4 GB estimatefull226 tok/s estimate
Q5 Q5 (~0.71 bytes/param)8,192 tokens5.7 GB estimate1.1 GB estimate0.8 GB estimate7.6 GB estimate31.4 GB estimatefull198 tok/s estimate
Q8 Q88,192 tokens8.6 GB estimate1.1 GB estimate0.8 GB estimate10.5 GB estimate31.4 GB estimatefull141 tok/s estimate
FP16 FP168,192 tokens16.1 GB estimate1.1 GB estimate0.8 GB estimate17.9 GB estimate31.4 GB estimatefull80 tok/s estimate
Q4, longer context32,768 tokens4.8 GB estimate4.3 GB estimate0.8 GB estimate9.9 GB estimate31.4 GB estimatefull149 tok/s estimate

Formula estimate, not a benchmark. Method →

Cheapest catalog option that fits at Q4

GeForce RTX 3060 12GB, used (median of sold listings), €257, dated 2026-10-16. No buy link.

This hardware

$7,052€6,299

new (lowest in-stock EU)

price as of 16 Oct 2026

PriceSquirrel

Cheapest in-stock listing on 16 Oct 2026: €6,299, ASUS ROG Astral GeForce RTX 5090 32GB OC White at Proshop (FI). MSI Gaming Trio OC at Compumsa (BE) was €6,571 and out of stock. A 16 Oct 2026 fact-check notes BE/NL/DE about €7,900–€8,800. No used median was recorded. NVIDIA (Jeff Fisher, 16 Oct 2026) denies a production halt; supply is tight and some retailers have delisted the card.

Embed this

Llama 3.1 8B on RTX 5090

[![Llama 3.1 8B Q4 on RTX 5090: fits ✓ · est. 226 tok/s (estimate)](https://localllmrig.com/badge/llama-3.1-8b/rtx-5090.svg)](https://localllmrig.com/can-i-run/llama-3.1-8b/rtx-5090?utm_source=badge)

Questions

Can I run Llama 3.1 8B on an RTX 5090?
Yes, fully in VRAM (estimate).
How much memory does Llama 3.1 8B need at Q4?
About 6.7 GB at 8,192 tokens (estimate): 4.8 GB weights, 1.1 GB KV cache, 0.8 GB overhead.
How fast is Llama 3.1 8B on RTX 5090 (estimate)?
About 226 tok/s at Q4 (estimate, not a benchmark).
Llama 3.1 8B on RTX 5090 (estimate) · localLLMrig