update

app.py

@@ -226,7 +226,12 @@ def create_leaderboard_ui():
             "- **likes**: Number of likes on HuggingFace Hub\n\n"
             "**WebGPU Compatibility:**\n"
             "- Models in the recommended section are all WebGPU compatible\n"
-            "- WebGPU enables GPU acceleration in modern browsers"
+            "- WebGPU enables GPU acceleration in modern browsers\n\n"
+            "**⚠️ Important Note About Performance Metrics:**\n"
+            "All metrics are measured in a controlled benchmark environment. "
+            "They are useful for **comparing models against each other**, but may not reflect "
+            "actual performance in your environment. Factors like hardware, browser, OS, and system load affect real-world performance. "
+            "We recommend testing models in your own environment for accurate measurements."
         )
 
         def update_data():

data_loader.py

@@ -745,7 +745,7 @@ def format_recommended_models_as_markdown(df: pd.DataFrame) -> str:
             infer_time = row.get("first_infer_ms_p50", None)
 
             # Model entry
-            markdown_lines.append(f"### {model_id}")
+            markdown_lines.append(f"#### {model_id}")
             markdown_lines.append("")
 
             # WebGPU compatibility
@@ -788,6 +788,19 @@ def format_recommended_models_as_markdown(df: pd.DataFrame) -> str:
         "- **Performance**: Fast loading and inference times based on benchmark results",
         "- **Compatibility**: Verified WebGPU support for GPU-accelerated browser execution",
         "",
+        "### Understanding Benchmark Metrics",
+        "",
+        "**Important:** All performance metrics (load time, inference time, etc.) are measured in a controlled benchmark environment. "
+        "These metrics are useful for **comparing models against each other**, but they may not reflect the actual performance you'll experience in your specific environment. "
+        "Factors that affect real-world performance include:",
+        "- Hardware specifications (CPU, GPU, memory)",
+        "- Browser type and version",
+        "- Operating system",
+        "- Network conditions (for model loading)",
+        "- Concurrent processes and system load",
+        "",
+        "**We recommend** benchmarking models in your own environment with your actual use case to get accurate performance measurements.",
+        "",
         "### For Production Use",
         "",
         "These recommendations are optimized for first-time trials and learning. "