merge

src/server/mod.rs

@@ -137,8 +137,23 @@ async fn get_model_cost(
     // Check in-memory cache for cost overrides (no SQLite hit)
     if let Some(cached) = state.model_config_cache.get(model).await {
         if let (Some(p), Some(c)) = (cached.prompt_cost_per_m, cached.completion_cost_per_m) {
-            // Manual overrides don't have cache-specific rates, so use simple formula
-            return (prompt_tokens as f64 * p / 1_000_000.0) + (completion_tokens as f64 * c / 1_000_000.0);
+            // Manual overrides logic: if cache rates are provided, use cache-aware formula.
+            // Formula: (non_cached_prompt * input_rate) + (cache_read * read_rate) + (cache_write * write_rate) + (completion * output_rate)
+            let non_cached_prompt = prompt_tokens.saturating_sub(cache_read_tokens);
+            let mut total = (non_cached_prompt as f64 * p / 1_000_000.0) + (completion_tokens as f64 * c / 1_000_000.0);
+            
+            if let Some(cr) = cached.cache_read_cost_per_m {
+                total += cache_read_tokens as f64 * cr / 1_000_000.0;
+            } else {
+                // No manual cache_read rate — charge cached tokens at full input rate (backwards compatibility)
+                total += cache_read_tokens as f64 * p / 1_000_000.0;
+            }
+            
+            if let Some(cw) = cached.cache_write_cost_per_m {
+                total += cache_write_tokens as f64 * cw / 1_000_000.0;
+            }
+            
+            return total;
         }
     }