RTX 4090 Qwen3-8B Q4 token-gen @4K
141.0 tok/s TG batch-1 @4K ctx
Fields
id
benchmarks/comm-rtx-4090-qwen3-8b-q4-tg-4kMetric
decode_tok_s
Value
141
Unit
tok/s TG batch-1 @4K ctx
Model
Qwen3-8B
Methodology
llama.cpp llama-bench, Q4 GGUF, CUDA, batch-1 decode @4K (Hardware Corner measured table)
Measured by
third_party
Thermal state
unknown
notes
14B 84 @4K; Qwen3-32B Q4 39.6 @4K / 34 @16K; gpt-oss-20B 190 @4K / 70 @128K; Qwen3-30B-A3B 207 @4K.
Notes
14B 84 @4K; Qwen3-32B Q4 39.6 @4K / 34 @16K; gpt-oss-20B 190 @4K / 70 @128K; Qwen3-30B-A3B 207 @4K.