NVIDIA GeForce RTX 3080 Ti — Benchmarks & Specs

Bottom line: how fast is the NVIDIA GeForce RTX 3080 Ti?

SpecPicks median: 89.7 tokens/sec across 7 cited single-user runs of 7-9B dense models at 4-bit (Q4) under llama.cpp (range 68.1–140 tok/s), from KnightLi llama.cpp GPU Scoreboard, Hardware Corner (3 runs), GPU-Benchmarks-on-LLM-Inference, LocalScore.ai (2 runs). At the $899.99 Amazon price of the cheapest linked listing (checked 2026-10-21 10:15 UTC; prices change), that is a SpecPicks Value Index of 99.7 tok/s per $1,000. At 1440p Ultra, the NVIDIA GeForce RTX 3080 Ti averages 58 fps in Death Stranding 2: On the Beach with a 53 fps 1% low, per PCBench. For local LLM inference it generates 115.2 tokens/sec running qwen3:8b at q4_K_M under ollama, per Smeltcore. In PassMark G3D Mark it scores 26,774 pts, per PassMark. Its 12 GB of VRAM is the binding constraint for local inference: the largest model on file at 4-bit quantization on this card is qwen3:14b (14B parameters, q4_K_M), generating 69.9 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.