Benchmark search

Try these searches

Found 41 run configurations

Each row is a run configuration (model + hardware + framework + quantization); repeated runs of the same setup are merged — raw counts appear in each result’s source line.

ModelQwen3.8-Flash-Next

Qwen3.8-Flash-Next · Q4_K_M

L0 Self-reported

NVIDIA RTX 5070 Ti · 16GB · Strata

Other

2026-10-06

8.2 tok/s

Decode

61.5 tok/s

Prefill

1.63 s

TTFT

16 GB

VRAM

Config (latest run)
Linux CUDA 12.6 Q4_K_M Strata 0.5.2 Flash Attention

Sources: Other · 1 runs · 1 independent sources

Qwen3.8-Flash-Next · IQ3_XXS

L0 Self-reported

NVIDIA RTX 5070 Ti · 16GB · Strata

Other

2026-10-06

14.2 tok/s

Decode

106.5 tok/s

Prefill

0.94 s

TTFT

16 GB

VRAM

Config (latest run)
Linux CUDA 12.6 IQ3_XXS Strata 0.5.2 Flash Attention

Sources: Other · 1 runs · 1 independent sources

Qwen3.8-Flash-Next · IQ3_S

L0 Self-reported

NVIDIA RTX 5070 Ti · 16GB · Strata

Other

2026-10-06

12.3 tok/s

Decode

92.3 tok/s

Prefill

1.08 s

TTFT

16 GB

VRAM

Config (latest run)
Linux CUDA 12.6 IQ3_S Strata 0.5.2 Flash Attention

Sources: Other · 1 runs · 1 independent sources

Qwen3.8-Flash-Next · IQ4_XS

L0 Self-reported

NVIDIA RTX 5070 Ti · 16GB · Strata

Other

2026-10-06

9.2 tok/s

Decode

69 tok/s

Prefill

1.45 s

TTFT

16 GB

VRAM

Config (latest run)
Linux CUDA 12.6 IQ4_XS Strata 0.5.2 Flash Attention

Sources: Other · 1 runs · 1 independent sources

Qwen3.8-Flash-Next · Q4_K_M

L0 Self-reported

NVIDIA RTX 5070 Ti · 16GB · llama.cpp

Other

2026-10-06

6.4 tok/s

Decode

48 tok/s

Prefill

2.08 s

TTFT

16 GB

VRAM

Config (latest run)
Linux CUDA 12.6 Q4_K_M llama.cpp b4200 Flash Attention

Sources: Other · 1 runs · 1 independent sources

Qwen3.8-Flash-Next · IQ3_XXS

L0 Self-reported

NVIDIA RTX 5070 Ti · 16GB · llama.cpp

Other

2026-10-06

11.1 tok/s

Decode

83.3 tok/s

Prefill

1.2 s

TTFT

16 GB

VRAM

Config (latest run)
Linux CUDA 12.6 IQ3_XXS llama.cpp b4200 Flash Attention

Sources: Other · 1 runs · 1 independent sources

Qwen3.8-Flash-Next · IQ4_XS

L0 Self-reported

NVIDIA RTX 5070 Ti · 16GB · llama.cpp

Other

2026-10-06

7.2 tok/s

Decode

54 tok/s

Prefill

1.85 s

TTFT

16 GB

VRAM

Config (latest run)
Linux CUDA 12.6 IQ4_XS llama.cpp b4200 Flash Attention

Sources: Other · 1 runs · 1 independent sources

Qwen3.8-Flash-Next · Q4_K_M

L0 Self-reported

NVIDIA RTX 3090 · 24GB · Strata

Other

2026-10-06

8.6 tok/s

Decode

64.5 tok/s

Prefill

1.55 s

TTFT

24 GB

VRAM

Config (latest run)
Linux CUDA 12.6 Q4_K_M Strata 0.5.2 Flash Attention

Sources: Other · 1 runs · 1 independent sources

Qwen3.8-Flash-Next · IQ3_XXS

L0 Self-reported

NVIDIA RTX 3090 · 24GB · Strata

Other

2026-10-06

14.8 tok/s

Decode

111 tok/s

Prefill

0.9 s

TTFT

24 GB

VRAM

Config (latest run)
Linux CUDA 12.6 IQ3_XXS Strata 0.5.2 Flash Attention

Sources: Other · 1 runs · 1 independent sources

Qwen3.8-Flash-Next · IQ3_S

L0 Self-reported

NVIDIA RTX 3090 · 24GB · Strata

Other

2026-10-06

12.8 tok/s

Decode

96 tok/s

Prefill

1.04 s

TTFT

24 GB

VRAM

Config (latest run)
Linux CUDA 12.6 IQ3_S Strata 0.5.2 Flash Attention

Sources: Other · 1 runs · 1 independent sources

Qwen3.8-Flash-Next · IQ4_XS

L0 Self-reported

NVIDIA RTX 3090 · 24GB · Strata

Other

2026-10-06

9.6 tok/s

Decode

72 tok/s

Prefill

1.39 s

TTFT

24 GB

VRAM

Config (latest run)
Linux CUDA 12.6 IQ4_XS Strata 0.5.2 Flash Attention

Sources: Other · 1 runs · 1 independent sources

Qwen3.8-Flash-Next · Q4_K_M

L0 Self-reported

NVIDIA RTX 3090 · 24GB · llama.cpp

Other

2026-10-06

6.7 tok/s

Decode

50.3 tok/s

Prefill

1.99 s

TTFT

24 GB

VRAM

Config (latest run)
Linux CUDA 12.6 Q4_K_M llama.cpp b4200 Flash Attention

Sources: Other · 1 runs · 1 independent sources

Qwen3.8-Flash-Next · IQ3_XXS

L0 Self-reported

NVIDIA RTX 3090 · 24GB · llama.cpp

Other

2026-10-06

11.6 tok/s

Decode

87 tok/s

Prefill

1.15 s

TTFT

24 GB

VRAM

Config (latest run)
Linux CUDA 12.6 IQ3_XXS llama.cpp b4200 Flash Attention

Sources: Other · 1 runs · 1 independent sources

Qwen3.8-Flash-Next · IQ4_XS

L0 Self-reported

NVIDIA RTX 3090 · 24GB · llama.cpp

Other

2026-10-06

7.5 tok/s

Decode

56.3 tok/s

Prefill

1.78 s

TTFT

24 GB

VRAM

Config (latest run)
Linux CUDA 12.6 IQ4_XS llama.cpp b4200 Flash Attention

Sources: Other · 1 runs · 1 independent sources

Qwen3.8-Flash-Next · Q4_K_M

L0 Self-reported

NVIDIA RTX 5090 · 32GB · Strata

Other

2026-10-06

16.4 tok/s

Decode

123 tok/s

Prefill

0.81 s

TTFT

32 GB

VRAM

Config (latest run)
Linux CUDA 12.6 Q4_K_M Strata 0.5.2 Flash Attention

Sources: Other · 1 runs · 1 independent sources

Qwen3.8-Flash-Next · IQ3_XXS

L0 Self-reported

NVIDIA RTX 5090 · 32GB · Strata

Other

2026-10-06

28.4 tok/s

Decode

213 tok/s

Prefill

0.47 s

TTFT

32 GB

VRAM

Config (latest run)
Linux CUDA 12.6 IQ3_XXS Strata 0.5.2 Flash Attention

Sources: Other · 1 runs · 1 independent sources

Qwen3.8-Flash-Next · IQ3_S

L0 Self-reported

NVIDIA RTX 5090 · 32GB · Strata

Other

2026-10-06

24.6 tok/s

Decode

184.5 tok/s

Prefill

0.54 s

TTFT

32 GB

VRAM

Config (latest run)
Linux CUDA 12.6 IQ3_S Strata 0.5.2 Flash Attention

Sources: Other · 1 runs · 1 independent sources

Qwen3.8-Flash-Next · IQ4_XS

L0 Self-reported

NVIDIA RTX 5090 · 32GB · Strata

Other

2026-10-06

18.4 tok/s

Decode

138 tok/s

Prefill

0.72 s

TTFT

32 GB

VRAM

Config (latest run)
Linux CUDA 12.6 IQ4_XS Strata 0.5.2 Flash Attention

Sources: Other · 1 runs · 1 independent sources

Qwen3.8-Flash-Next · Q4_K_M

L0 Self-reported

NVIDIA RTX 5090 · 32GB · llama.cpp

Other

2026-10-06

12.8 tok/s

Decode

96 tok/s

Prefill

1.04 s

TTFT

32 GB

VRAM

Config (latest run)
Linux CUDA 12.6 Q4_K_M llama.cpp b4200 Flash Attention

Sources: Other · 1 runs · 1 independent sources

Qwen3.8-Flash-Next · IQ3_XXS

L0 Self-reported

NVIDIA RTX 5090 · 32GB · llama.cpp

Other

2026-10-06

22.1 tok/s

Decode

165.8 tok/s

Prefill

0.6 s

TTFT

32 GB

VRAM

Config (latest run)
Linux CUDA 12.6 IQ3_XXS llama.cpp b4200 Flash Attention

Sources: Other · 1 runs · 1 independent sources

41 results

Related models
Related hardware
Related frameworks
Related quantizations