diff --git a/benchmarks/Meta-Llama-3.1-8B-Instruct/A10G/20-08-2024.json b/benchmarks/Meta-Llama-3.1-8B-Instruct/A10G/20-08-2024.json
new file mode 100644
index 0000000..ef30f8d
--- /dev/null
+++ b/benchmarks/Meta-Llama-3.1-8B-Instruct/A10G/20-08-2024.json
@@ -0,0 +1,18 @@
+{
+    "date": "20240820-143000", 
+    "provider": "aws",
+    "gpu": "NVIDIA A10G",
+    "cost_per_hour": 2.84,
+    "backend": "vllm", 
+    "model_id": "meta-llama/Meta-Llama-3.1-8B-Instruct", 
+    "tokenizer_id": "meta-llama/Meta-Llama-3.1-8B-Instruct", 
+    "num_prompts": 1000, 
+    "duration": 97.65, 
+    "completed": 1000, 
+    "total_input_tokens": 215196, 
+    "total_output_tokens": 43140, 
+    "request_throughput": 10.24, 
+    "input_throughput": 2203.64, 
+    "output_throughput": 441.76,
+    "issue": "https://github.com/arc53/llm-price-compass/issues/6"
+}
\ No newline at end of file