NVIDIA GeForce RTX 4070 Ti SUPER — Benchmarks & Specs

Bottom line: how fast is the NVIDIA GeForce RTX 4070 Ti SUPER?

SpecPicks median: 75.4 tokens/sec across 4 cited single-user runs of 7-9B dense models at 4-bit (Q4) under llama.cpp (range 36.6–132 tok/s), from llama.cpp GitHub, Hardware Corner (3 runs). At 1440p Ultra with DLSS Quality upscaling, the NVIDIA GeForce RTX 4070 Ti SUPER averages 113 fps in Cyberpunk 2077 with a 57 fps 1% low, per WorthPlaying. For local LLM inference it generates 36.8 tokens/sec running qwen2.5:14b at q4_K_M under llamafile, per LocalScore. In 3DMark Fire Strike it scores 42,724 points, per WorthPlaying. 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, MXFP4), generating 128.9 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. SpecPicks aggregates published measurements; it does not report first-party benchmark runs.