AMD Radeon RX 6600 XT — Benchmarks & Specs

Bottom line: how fast is the Radeon RX 6600 XT?

SpecPicks median: 34.0 tokens/sec across 5 cited single-user runs of 7-9B dense models at 4-bit (Q4) under llama.cpp (range 20.4–64.5 tok/s), from knightli.com, llama.cpp GitHub, GitHub: ChharithOeun/llm-amd-windows, DEV Community (2 runs). At the $498.00 Amazon price of the cheapest linked listing (checked 2026-10-13 22:14 UTC; prices change), that is 68.3 tok/s per $1,000. At 1440p Ultra, the Radeon RX 6600 XT averages 69 fps in Horizon Zero Dawn, per Trusted Reviews. For local LLM inference it generates 20.9 tokens/sec running llama3.1:8b at q4_K_M under llama.cpp, per DEV Community. In PassMark G3D Mark it scores 16,450 pts, per PassMark. Its 8 GB of VRAM is the binding constraint for local inference: the llama3.1:8b run above (8B parameters) is the largest model on file at 4-bit quantization on this card.

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.