NVIDIA RTX 5070 llama.cpp: Qwen3 14B (Q4_K) prompt processing at 4k context
2381.6 tok/s (pp4k, per-request batch-1)
Fields
id
benchmarks/hc26-5070-qwen3-14b-q4k-4k-ppMetric
prefill_tok_s
Value
2,381.6
Unit
tok/s (pp4k, per-request batch-1)
Model
Qwen3 14B (Q4_K)
Methodology
THIRD-PARTY MEASURED (Hardware Corner GPU LLM Benchmarks page, tables dated March 2026). llama.cpp llama-bench, GGUF, single GPU, batch-1. NVIDIA RTX 5070. Qwen3 14B (Q4_K). Number is the 4k-context column of the Prompt Processing table: context depth the run was measured at, not a fixed pp512/tg128. Per-request, NOT aggregate.
Measured by
third_party
Reproducible
yes