feat: implement web UI for provider and model configuration

- Added 'provider_configs' and 'model_configs' tables to database.
- Refactored ProviderManager to support thread-safe dynamic updates and database overrides.
- Implemented 'Models' tab in dashboard to manage model visibility, mapping, and pricing.
- Added provider configuration modal to 'Providers' tab.
- Integrated database overrides into chat completion logic (enabled state, mapping, and cost).

src/utils/streaming.rs

@@ -2,6 +2,7 @@ use futures::stream::Stream;
 use std::pin::Pin;
 use std::task::{Context, Poll};
 use std::sync::Arc;
+use sqlx::Row;
 use crate::logging::{RequestLogger, RequestLog};
 use crate::client::ClientManager;
 use crate::providers::{Provider, ProviderStreamChunk};
@@ -20,6 +21,7 @@ pub struct AggregatingStream<S> {
     logger: Arc<RequestLogger>,
     client_manager: Arc<ClientManager>,
     model_registry: Arc<crate::models::registry::ModelRegistry>,
+    db_pool: crate::database::DbPool,
     start_time: std::time::Instant,
     has_logged: bool,
 }
@@ -38,6 +40,7 @@ where
         logger: Arc<RequestLogger>,
         client_manager: Arc<ClientManager>,
         model_registry: Arc<crate::models::registry::ModelRegistry>,
+        db_pool: crate::database::DbPool,
     ) -> Self {
         Self {
             inner,
@@ -51,6 +54,7 @@ where
             logger,
             client_manager,
             model_registry,
+            db_pool,
             start_time: std::time::Instant::now(),
             has_logged: false,
         }
@@ -72,6 +76,7 @@ where
         let prompt_tokens = self.prompt_tokens;
         let has_images = self.has_images;
         let registry = self.model_registry.clone();
+        let pool = self.db_pool.clone();
         
         // Estimate completion tokens (including reasoning if present)
         let content_tokens = estimate_completion_tokens(&self.accumulated_content, &model);
@@ -83,10 +88,29 @@ where
         
         let completion_tokens = content_tokens + reasoning_tokens;
         let total_tokens = prompt_tokens + completion_tokens;
-        let cost = provider.calculate_cost(&model, prompt_tokens, completion_tokens, &registry);
 
         // Spawn a background task to log the completion
         tokio::spawn(async move {
+            // Check database for cost overrides
+            let db_cost = sqlx::query("SELECT prompt_cost_per_m, completion_cost_per_m FROM model_configs WHERE id = ?")
+                .bind(&model)
+                .fetch_optional(&pool)
+                .await
+                .unwrap_or(None);
+
+            let cost = if let Some(row) = db_cost {
+                let prompt_rate = row.get::<Option<f64>, _>("prompt_cost_per_m");
+                let completion_rate = row.get::<Option<f64>, _>("completion_cost_per_m");
+
+                if let (Some(p), Some(c)) = (prompt_rate, completion_rate) {
+                    (prompt_tokens as f64 * p / 1_000_000.0) + (completion_tokens as f64 * c / 1_000_000.0)
+                } else {
+                    provider.calculate_cost(&model, prompt_tokens, completion_tokens, &registry)
+                }
+            } else {
+                provider.calculate_cost(&model, prompt_tokens, completion_tokens, &registry)
+            };
+
             // Log to database
             logger.log_request(RequestLog {
                 timestamp: chrono::Utc::now(),
@@ -188,6 +212,7 @@ mod tests {
             logger,
             client_manager,
             registry,
+            pool.clone(),
         );
         
         while let Some(item) = agg_stream.next().await {