TinyLlama on NVIDIA GeForce RTX 5090 · 32GB VRAM

✅ Runs comfortably — TinyLlama 1.1B @ FP16
TinyLlama 1.1B at FP16 needs ~2.7 GB (weights 2.0 GB + KV 176 MB + overhead 512 MB @ 8K ctx) of your 28.8 GB usable VRAM — plenty of headroom. Expect ~330–540 tok/s (very fast).
context @ FP16: comfortable to 2K
quantneedsspeed
✓ FP162.7 GB~330–540 tok/s (very fast)
✓ Q8_01.8 GB~580–960 tok/s (very fast)
✓ Q4_K_M1.3 GB~920–1530 tok/s (very fast)
$runlocal install tinyllama:1.1b
⚠ NVIDIA support is best-effort in v0.1 — verify before relying on it.
runlocal verdict card
download card share on x share on reddit drop it in your model-drop post
TinyLlama: Runs comfortably
for your own README — links back here

Same model, other machines: Apple M1 · Apple M1 Pro · Apple M2 · Apple M3 Pro · Apple M2 Pro · Apple M4 Pro

Also runs on a NVIDIA GeForce RTX 5090: DeepSeek-R1 (distill) · Qwen2.5 · Qwen3 · Llama 3.1 · Gemma 3

Best models for a NVIDIA GeForce RTX 5090 · 32GB VRAM →

Check any combo yourself: open the checker →