AMD Radeon RX 6800 — Benchmarks & Specs

Bottom line: how fast is the Radeon RX 6800?

SpecPicks median: 83.9 tokens/sec across 3 cited single-user runs of 7-9B dense models at 4-bit (Q4) under llama.cpp (range 48.0–95.6 tok/s), from llama.cpp GitHub Discussions (2 runs), GigaChad LLC. At the $489.97 Amazon price of the cheapest linked listing (checked 2026-10-13 16:16 UTC; prices change), that is 171.3 tok/s per $1,000. At 1440p Ultra, the Radeon RX 6800 averages 68 fps in Alan Wake 2 with a 52 fps 1% low, per Guru3D. For local LLM inference it generates 95.6 tokens/sec running llama2:7b at Q4_0 under llama.cpp, per llama.cpp GitHub Discussions. In PassMark G3D Mark it scores 22,023 pts, per PassMark. It has 16 GB of VRAM, and the largest model on file at 4-bit quantization on this card is qwen:14b (14B parameters, q4_K_M), generating 24.0 tokens/sec under llama.cpp, per GigaChad LLC. Larger models have no sourced run on file.

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.