Grouped by reported model identity. Exact weights, quantizations and software can differ. How results are reviewed
Documented results
24 selected tests| View Reportllama-bench | gemma-3-27b-it | Token speed512 input tokens · 0 existing KV tokens | — | 568.31 tok/sPrefill · mean · per request | — | 2 × R9700Multi-GPU · split unknown · 32 GB VRAM/GPUConfiguration
|
|---|---|---|---|---|---|---|
| View ReportGeneric LLM | gemma-3-27b-it | Token speed512 input tokens · 0 existing KV tokens | — | 659.05 tok/sPrefill · mean · per request | — | 1 × R9700Single GPU · 32 GB VRAM/GPUConfiguration
|
| View Reportllama-bench | gemma-3-27b-it | Token speed128 output tokens · 0 existing KV tokens | — | — | 9.91 tok/sGeneration · mean · per request | 2 × R9700Multi-GPU · split unknown · 32 GB VRAM/GPUConfiguration
|
| View ReportGeneric LLM | gemma-3-27b-it | Token speed128 output tokens · 0 existing KV tokens | — | — | 23.25 tok/sGeneration · mean · per request | 1 × R9700Single GPU · 32 GB VRAM/GPUConfiguration
|
| View Reportllama-bench | gemma-3-27b-it | Token speed2,048 input tokens · 16,384 existing KV tokens | — | 654.31 tok/sPrefill · mean · per request | — | 2 × R9700Multi-GPU · split unknown · 32 GB VRAM/GPUConfiguration
|
| View ReportGeneric LLM | gemma-3-27b-it | Token speed512 input tokens · 4,000 existing KV tokens | — | 582.29 tok/sPrefill · mean · per request | — | 1 × R9700Single GPU · 32 GB VRAM/GPUConfiguration
|
| View Reportllama-bench | gemma-3-27b-it | Token speed32 output tokens · 16,384 existing KV tokens | — | — | 9.48 tok/sGeneration · mean · per request | 2 × R9700Multi-GPU · split unknown · 32 GB VRAM/GPUConfiguration
|
| View ReportGeneric LLM | gemma-3-27b-it | Token speed128 output tokens · 4,000 existing KV tokens | — | — | 21.04 tok/sGeneration · mean · per request | 1 × R9700Single GPU · 32 GB VRAM/GPUConfiguration
|
| View Reportllama-bench | gemma-3-27b-it | Token speed2,048 input tokens · 32,768 existing KV tokens | — | 528.1 tok/sPrefill · mean · per request | — | 2 × R9700Multi-GPU · split unknown · 32 GB VRAM/GPUConfiguration
|
| View ReportGeneric LLM | gemma-3-27b-it | Token speed512 input tokens · 8,000 existing KV tokens | — | 531.76 tok/sPrefill · mean · per request | — | 1 × R9700Single GPU · 32 GB VRAM/GPUConfiguration
|
| View Reportllama-bench | gemma-3-27b-it | Token speed32 output tokens · 32,768 existing KV tokens | — | — | 9.12 tok/sGeneration · mean · per request | 2 × R9700Multi-GPU · split unknown · 32 GB VRAM/GPUConfiguration
|
| View ReportGeneric LLM | gemma-3-27b-it | Token speed128 output tokens · 8,000 existing KV tokens | — | — | 22.2 tok/sGeneration · mean · per request | 1 × R9700Single GPU · 32 GB VRAM/GPUConfiguration
|
| View Reportllama-bench | gemma-3-27b-it | Token speed512 input tokens · 0 existing KV tokens | — | 1,493.5 tok/sPrefill · mean · per request | — | 2 × R9700Multi-GPU · split unknown · 32 GB VRAM/GPUConfiguration
|
| View ReportGeneric LLM | gemma-3-27b-it | Token speed512 input tokens · 16,000 existing KV tokens | — | 478.3 tok/sPrefill · mean · per request | — | 1 × R9700Single GPU · 32 GB VRAM/GPUConfiguration
|
| View Reportllama-bench | gemma-3-27b-it | Token speed128 output tokens · 0 existing KV tokens | — | — | 10.61 tok/sGeneration · mean · per request | 2 × R9700Multi-GPU · split unknown · 32 GB VRAM/GPUConfiguration
|
| View ReportGeneric LLM | gemma-3-27b-it | Token speed128 output tokens · 16,000 existing KV tokens | — | — | 21.67 tok/sGeneration · mean · per request | 1 × R9700Single GPU · 32 GB VRAM/GPUConfiguration
|
| View Reportllama-bench | gemma-3-27b-it | Token speed2,048 input tokens · 16,384 existing KV tokens | — | 941.08 tok/sPrefill · mean · per request | — | 2 × R9700Multi-GPU · split unknown · 32 GB VRAM/GPUConfiguration
|
| View ReportGeneric LLM | gemma-3-27b-it | Token speed512 input tokens · 32,000 existing KV tokens | — | 418.48 tok/sPrefill · mean · per request | — | 1 × R9700Single GPU · 32 GB VRAM/GPUConfiguration
|
| View Reportllama-bench | gemma-3-27b-it | Token speed32 output tokens · 16,384 existing KV tokens | — | — | 10.31 tok/sGeneration · mean · per request | 2 × R9700Multi-GPU · split unknown · 32 GB VRAM/GPUConfiguration
|
| View ReportGeneric LLM | gemma-3-27b-it | Token speed128 output tokens · 32,000 existing KV tokens | — | — | 20.71 tok/sGeneration · mean · per request | 1 × R9700Single GPU · 32 GB VRAM/GPUConfiguration
|
| View Reportllama-bench | gemma-3-27b-it | Token speed512 input tokens · 0 existing KV tokens | — | 180.71 tok/sPrefill · mean · per request | — | 2 × R9700Multi-GPU · split unknown · 32 GB VRAM/GPUConfiguration
|
| View ReportGeneric LLM | gemma-3-27b-it | Token speed512 input tokens · 48,000 existing KV tokens | — | 373.22 tok/sPrefill · mean · per request | — | 1 × R9700Single GPU · 32 GB VRAM/GPUConfiguration
|
| View Reportllama-bench | gemma-3-27b-it | Token speed128 output tokens · 0 existing KV tokens | — | — | 10.45 tok/sGeneration · mean · per request | 2 × R9700Multi-GPU · split unknown · 32 GB VRAM/GPUConfiguration
|
| View ReportGeneric LLM | gemma-3-27b-it | Token speed128 output tokens · 48,000 existing KV tokens | — | — | 19.78 tok/sGeneration · mean · per request | 1 × R9700Single GPU · 32 GB VRAM/GPUConfiguration
|
Quantizations and software
Quantizations: BF16, Q6_K_L
Engines: llama.cpp
This is a selection across sources, not a ranking or a matched experiment. Reading benchmark reports
GPU benchmark pages: R9700