NVIDIA GeForce RTX 5080 — Benchmarks & Specs

Bottom line: how fast is the NVIDIA GeForce RTX 5080?

SpecPicks median: 138 tokens/sec across 4 cited single-user runs of 7-9B dense models at 4-bit (Q4) under llama.cpp (range 44.9–185 tok/s), from llama.cpp GitHub (2 runs), Hardware Corner, LocalScore. At the $2,099.99 Amazon price of the cheapest linked listing (checked 2026-10-21 12:20 UTC; prices change), that is a SpecPicks Value Index of 65.7 tok/s per $1,000. At 1440p Ultra, the NVIDIA GeForce RTX 5080 averages 131 fps in Cyberpunk 2077, per KitGuru. For local LLM inference it generates 25.5 tokens/sec running Qwen2.5 14B Instruct at Q4_K_M under llama.cpp, per LocalScore. In 3DMark Fire Strike it scores 79,045 points, per LanOC Reviews. Its 16 GB of VRAM is the binding constraint for local inference: the largest model on file at 4-bit quantization on this card is gpt-oss:20b (20B parameters, q4_K_XL), generating 140.5 tokens/sec under llama.cpp, per Hardware Corner.

Every figure above is a row in the tables below, and each row links out to the review or public benchmark database the number was taken from; the SpecPicks median is computed here from the runs it names, each a row in the AI table, and the per-$1,000 figure divides it by the dated listing price. SpecPicks aggregates published measurements; it does not report first-party benchmark runs.