Feat/usage improvements (#508)

* i18n: update cache terminology across all languages

- Change 'Cache Read' to 'Cache Hit' in all languages
- Change 'Cache Write' to 'Cache Creation' in all languages
- Update zh: 缓存读取 → 缓存命中, 缓存写入 → 缓存创建
- Update en: Cache Read → Cache Hit, Cache Write → Cache Creation
- Update ja: キャッシュ読取 → キャッシュヒット, キャッシュ書込 → キャッシュ作成

Affected keys: cacheReadTokens, cacheCreationTokens, cacheReadCost,
cacheWriteCost, cacheRead, cacheWrite

* feat(usage): add cache metrics to trend chart

- Add cache creation tokens visualization (orange line)
- Add cache hit tokens visualization (purple line)
- Add gradient definitions for new cache metrics
- Include cache data in hourly aggregation
- Display cache metrics alongside input/output tokens

This provides better visibility into cache usage patterns over time.

* fix(usage): fix timezone handling in datetime picker

- Add timestampToLocalDatetime() to convert Unix timestamp to local datetime
- Add localDatetimeToTimestamp() with validation for incomplete input
- Fix issue where typing hours/minutes would jump to previous day
- Validate datetime format completeness before conversion
- Use local timezone instead of UTC for datetime-local input

This resolves the issue where users couldn't fine-tune time selection
and the input would jump unexpectedly when editing hours or minutes.

* feat(usage): add auto-refresh for usage statistics

- Add 30-second auto-refresh interval for all usage queries
- Disable background refresh to save resources
- Apply to: summary, trends, provider stats, model stats, request logs
- Queries automatically update when tab is active
- Pause refresh when user switches to another tab

This keeps usage data fresh without manual refresh.

* fix(proxy): improve usage logging and cache token parsing

- Log requests even when usage parsing fails (with default values)
- Add detailed debug logging for usage metrics
- Support cache_read_input_tokens field in Codex responses
- Fallback to input_tokens_details.cached_tokens if needed
- Add test case for cached_tokens in input_tokens_details
- Ensure all requests are tracked in database for analytics

This fixes missing request logs when API responses lack usage data
and improves cache token detection across different response formats.

* style(rust): use inline format args in format! macros

- Replace format!("...", var) with format!("...{var}")
- Update universal provider ID formatting
- Update error message formatting
- Update config.toml generation in Codex provider

Fixes clippy::uninlined_format_args warnings.

* feat(proxy): enhance provider router logging

- Add debug logs for failover queue provider count
- Log circuit breaker state for each provider check
- Add logs for missing current provider scenarios
- Log when no current provider is configured
- Use inline format args for better readability

This improves debugging of provider selection and failover behavior.

* feat(database): update model pricing data

- Update Claude models to full version format (e.g. claude-opus-4-5-20251101)
- Add GPT-5.2 series model pricing (10 models)
- Add GPT-5.1 series model pricing (10 models)
- Add GPT-5 series model pricing (12 models)
- Add Gemini 3 series model pricing (2 models)
- Update Gemini 2.5 series model ID format (use dot separator)
- Unify display names by removing thinking level suffixes

* fix(usage): correct Gemini output token calculation

Fix Gemini API output token parsing to use totalTokenCount - promptTokenCount
instead of candidatesTokenCount alone. This ensures thoughtsTokenCount is
included in output statistics.

- Update from_gemini_response to calculate output from total - input
- Update from_gemini_stream_chunks with same logic for consistency
- Fix from_codex_stream_events to use adjusted token calculation
- Add test case for responses with thoughtsTokenCount
- Update existing tests to match new calculation logic

* fix(usage): correct cache token billing and add Codex format auto-detection

- Avoid double-billing cache tokens by subtracting from input before calculation
- Add smart Codex parser that auto-detects OpenAI vs Codex API format
- Extract model name from Codex responses for accurate tracking

* fix(proxy): improve takeover detection with live config check

- Add live config takeover detection for hot-switch decision
- Rebuild takeover when backup is missing or placeholder remains
- Make detect_takeover_in_live_config_for_app public
- Fix is_takeover_active to use actual takeover status

* refactor(usage): simplify model pricing lookup by removing suffix fallback

Replace complex suffix-stripping fallback with direct prefix/suffix cleanup.
Model IDs are now cleaned by removing vendor prefix (before /) and colon
suffix (after :), then matched exactly against pricing table.

* feat(database): add Chinese AI model pricing data

Add pricing for domestic AI models (CNY/1M tokens):
- Doubao-Seed-Code (ByteDance)
- DeepSeek V3/V3.1/V3.2
- Kimi K2/K2-Thinking/K2-Turbo (Moonshot)
- MiniMax M2/M2.1/M2.1-Lightning
- GLM-4.6/4.7 (Zhipu)
- Mimo V2 Flash (Xiaomi)

Also fix test case to use correct model ID and remove invalid currency column.

* refactor(proxy): improve header forwarding with blacklist approach

Change from whitelist to blacklist mode for request header forwarding.
Only skip headers that will be overridden (auth, host, content-length).
This preserves client's original headers and improves compatibility.

* fix(proxy): bypass timeout and retry configs when failover is disabled

When auto_failover_enabled is false, timeout and retry configurations
should not affect normal request flow. This change ensures:

- create_forwarder: passes 0 for all timeout/retry params when failover
  is disabled, effectively bypassing these checks
- streaming_timeout_config: returns 0 for both first_byte_timeout and
  idle_timeout when failover is disabled

This prevents unnecessary timeout errors and retry attempts when users
have explicitly disabled the failover feature.

* fix(proxy): handle zero value input in failover config fields

* refactor(proxy): remove retry logic and add enabled check for failover

* refactor(proxy): distinguish circuit-open from no-provider errors

* Align usage stats to sliding windows

* feat(proxy): add body and header filtering for upstream requests

* feat(proxy): enable transparent passthrough for headers

- Passthrough anthropic-beta header as-is from client
- Passthrough anthropic-version header from client
- Passthrough client IP headers (x-forwarded-for, x-real-ip) by default
- Filter private params (underscore-prefixed fields) from request body
- No database changes required

* feat(proxy): extract session ID from client requests for logging

- Add SessionIdExtractor to parse session ID from Claude/Codex requests
- Support extraction from metadata.user_id, headers, previous_response_id
- Pass session_id through RequestContext to usage logger
- Enable request correlation by session in proxy_request_logs

src-tauri/src/services/provider/mod.rs

@@ -217,9 +217,12 @@ impl ProviderService {
                 .flatten()
                 .is_some();
         let is_proxy_running = futures::executor::block_on(state.proxy_service.is_running());
+        let live_taken_over = state
+            .proxy_service
+            .detect_takeover_in_live_config_for_app(&app_type);
 
         // Hot-switch only when BOTH: this app is taken over AND proxy server is actually running
-        let should_hot_switch = is_app_taken_over && is_proxy_running;
+        let should_hot_switch = (is_app_taken_over || live_taken_over) && is_proxy_running;
 
         if should_hot_switch {
             // Proxy takeover mode: hot-switch only, don't write Live config
@@ -736,15 +739,15 @@ impl ProviderService {
         // 删除生成的子供应商
         if let Some(p) = provider {
             if p.apps.claude {
-                let claude_id = format!("universal-claude-{}", id);
+                let claude_id = format!("universal-claude-{id}");
                 let _ = state.db.delete_provider("claude", &claude_id);
             }
             if p.apps.codex {
-                let codex_id = format!("universal-codex-{}", id);
+                let codex_id = format!("universal-codex-{id}");
                 let _ = state.db.delete_provider("codex", &codex_id);
             }
             if p.apps.gemini {
-                let gemini_id = format!("universal-gemini-{}", id);
+                let gemini_id = format!("universal-gemini-{id}");
                 let _ = state.db.delete_provider("gemini", &gemini_id);
             }
         }
@@ -757,7 +760,7 @@ impl ProviderService {
         let provider = state
             .db
             .get_universal_provider(id)?
-            .ok_or_else(|| AppError::Message(format!("统一供应商 {} 不存在", id)))?;
+            .ok_or_else(|| AppError::Message(format!("统一供应商 {id} 不存在")))?;
 
         // 同步到 Claude
         if let Some(mut claude_provider) = provider.to_claude_provider() {
@@ -770,7 +773,7 @@ impl ProviderService {
             state.db.save_provider("claude", &claude_provider)?;
         } else {
             // 如果禁用了 Claude，删除对应的子供应商
-            let claude_id = format!("universal-claude-{}", id);
+            let claude_id = format!("universal-claude-{id}");
             let _ = state.db.delete_provider("claude", &claude_id);
         }
 
@@ -784,7 +787,7 @@ impl ProviderService {
             }
             state.db.save_provider("codex", &codex_provider)?;
         } else {
-            let codex_id = format!("universal-codex-{}", id);
+            let codex_id = format!("universal-codex-{id}");
             let _ = state.db.delete_provider("codex", &codex_id);
         }
 
@@ -798,7 +801,7 @@ impl ProviderService {
             }
             state.db.save_provider("gemini", &gemini_provider)?;
         } else {
-            let gemini_id = format!("universal-gemini-{}", id);
+            let gemini_id = format!("universal-gemini-{id}");
             let _ = state.db.delete_provider("gemini", &gemini_id);
         }
 

src-tauri/src/services/proxy.rs

@@ -193,7 +193,7 @@ impl ProxyService {
                 self.start().await?;
             }
 
-            // 2) 已接管则直接返回（幂等）
+            // 2) 已接管则直接返回（幂等）；但如果缺少备份或占位符残留，需要重建接管
             let current_config = self
                 .db
                 .get_proxy_config_for_app(app_type_str)
@@ -201,7 +201,22 @@ impl ProxyService {
                 .map_err(|e| format!("获取 {app_type_str} 配置失败: {e}"))?;
 
             if current_config.enabled {
-                return Ok(());
+                let has_backup = match self.db.get_live_backup(app_type_str).await {
+                    Ok(v) => v.is_some(),
+                    Err(e) => {
+                        log::warn!("读取 {app_type_str} 备份失败（将继续重建接管）: {e}");
+                        false
+                    }
+                };
+                let live_taken_over = self.detect_takeover_in_live_config_for_app(&app);
+
+                if has_backup || live_taken_over {
+                    return Ok(());
+                }
+
+                log::warn!(
+                    "{app_type_str} 标记为已接管，但缺少备份或占位符，正在重新接管并补齐备份"
+                );
             }
 
             // 3) 备份 Live 配置（严格：目标 app 不存在则报错）
@@ -1063,7 +1078,7 @@ impl ProxyService {
         }
     }
 
-    fn detect_takeover_in_live_config_for_app(&self, app_type: &AppType) -> bool {
+    pub fn detect_takeover_in_live_config_for_app(&self, app_type: &AppType) -> bool {
         match app_type {
             AppType::Claude => match self.read_claude_live() {
                 Ok(config) => Self::is_claude_live_taken_over(&config),
@@ -1257,10 +1272,8 @@ impl ProxyService {
 
     /// 检查是否处于 Live 接管模式
     pub async fn is_takeover_active(&self) -> Result<bool, String> {
-        self.db
-            .is_live_takeover_active()
-            .await
-            .map_err(|e| format!("检查接管状态失败: {e}"))
+        let status = self.get_takeover_status().await?;
+        Ok(status.claude || status.codex || status.gemini)
     }
 
     /// 从异常退出中恢复（启动时调用）

src-tauri/src/services/usage_stats.rs

@@ -4,7 +4,7 @@
 
 use crate::database::{lock_conn, Database};
 use crate::error::AppError;
-use chrono::{Duration, Local, TimeZone};
+use chrono::{Local, TimeZone};
 use rusqlite::{params, Connection, OptionalExtension};
 use serde::{Deserialize, Serialize};
 use serde_json::Value;
@@ -181,145 +181,114 @@ impl Database {
         Ok(result)
     }
 
-    /// 获取每日趋势
-    pub fn get_daily_trends(&self, days: u32) -> Result<Vec<DailyStats>, AppError> {
+    /// 获取每日趋势（滑动窗口，<=24h 按小时，>24h 按天，窗口与汇总一致）
+    pub fn get_daily_trends(
+        &self,
+        start_date: Option<i64>,
+        end_date: Option<i64>,
+    ) -> Result<Vec<DailyStats>, AppError> {
         let conn = lock_conn!(self.conn);
 
-        if days <= 1 {
-            let today = Local::now().date_naive();
-            let start_of_today = today.and_hms_opt(0, 0, 0).unwrap();
-            // 使用 earliest() 处理 DST 切换时的歧义时间，fallback 到当前时间减一天
-            let start_ts = Local
-                .from_local_datetime(&start_of_today)
-                .earliest()
-                .unwrap_or_else(|| Local::now() - Duration::days(1))
-                .timestamp();
+        let end_ts = end_date.unwrap_or_else(|| Local::now().timestamp());
+        let mut start_ts = start_date.unwrap_or_else(|| end_ts - 24 * 60 * 60);
 
-            let sql = "SELECT 
-                    strftime('%Y-%m-%dT%H:00:00', datetime(created_at, 'unixepoch', 'localtime')) as bucket,
-                    COUNT(*) as request_count,
-                    COALESCE(SUM(CAST(total_cost_usd AS REAL)), 0) as total_cost,
-                    COALESCE(SUM(input_tokens + output_tokens), 0) as total_tokens,
-                    COALESCE(SUM(input_tokens), 0) as total_input_tokens,
-                    COALESCE(SUM(output_tokens), 0) as total_output_tokens,
-                    COALESCE(SUM(cache_creation_tokens), 0) as total_cache_creation_tokens,
-                    COALESCE(SUM(cache_read_tokens), 0) as total_cache_read_tokens
-                 FROM proxy_request_logs
-                 WHERE created_at >= ?
-                 GROUP BY bucket
-                 ORDER BY bucket ASC";
-
-            let mut stmt = conn.prepare(sql)?;
-            let rows = stmt.query_map([start_ts], |row| {
-                Ok(DailyStats {
-                    date: row.get(0)?,
-                    request_count: row.get::<_, i64>(1)? as u64,
-                    total_cost: format!("{:.6}", row.get::<_, f64>(2)?),
-                    total_tokens: row.get::<_, i64>(3)? as u64,
-                    total_input_tokens: row.get::<_, i64>(4)? as u64,
-                    total_output_tokens: row.get::<_, i64>(5)? as u64,
-                    total_cache_creation_tokens: row.get::<_, i64>(6)? as u64,
-                    total_cache_read_tokens: row.get::<_, i64>(7)? as u64,
-                })
-            })?;
-
-            let mut buckets: HashMap<String, DailyStats> = HashMap::new();
-            for row in rows {
-                let stat = row?;
-                buckets.insert(stat.date.clone(), stat);
-            }
-
-            let mut stats = Vec::new();
-            for hour in 0..24 {
-                let bucket = today
-                    .and_hms_opt(hour, 0, 0)
-                    .unwrap()
-                    .format("%Y-%m-%dT%H:00:00")
-                    .to_string();
-
-                if let Some(stat) = buckets.remove(&bucket) {
-                    stats.push(stat);
-                } else {
-                    stats.push(DailyStats {
-                        date: bucket,
-                        request_count: 0,
-                        total_cost: "0.000000".to_string(),
-                        total_tokens: 0,
-                        total_input_tokens: 0,
-                        total_output_tokens: 0,
-                        total_cache_creation_tokens: 0,
-                        total_cache_read_tokens: 0,
-                    });
-                }
-            }
-            Ok(stats)
-        } else {
-            let today = Local::now().date_naive();
-            let start_day = today - Duration::days((days.saturating_sub(1)) as i64);
-            let start_of_window = start_day.and_hms_opt(0, 0, 0).unwrap();
-            // 使用 earliest() 处理 DST 切换时的歧义时间，fallback 到当前时间减 days 天
-            let start_ts = Local
-                .from_local_datetime(&start_of_window)
-                .earliest()
-                .unwrap_or_else(|| Local::now() - Duration::days(days as i64))
-                .timestamp();
-
-            let sql = "SELECT 
-                    strftime('%Y-%m-%dT00:00:00', datetime(created_at, 'unixepoch', 'localtime')) as bucket,
-                    COUNT(*) as request_count,
-                    COALESCE(SUM(CAST(total_cost_usd AS REAL)), 0) as total_cost,
-                    COALESCE(SUM(input_tokens + output_tokens), 0) as total_tokens,
-                    COALESCE(SUM(input_tokens), 0) as total_input_tokens,
-                    COALESCE(SUM(output_tokens), 0) as total_output_tokens,
-                    COALESCE(SUM(cache_creation_tokens), 0) as total_cache_creation_tokens,
-                    COALESCE(SUM(cache_read_tokens), 0) as total_cache_read_tokens
-                 FROM proxy_request_logs
-                 WHERE created_at >= ?
-                 GROUP BY bucket
-                 ORDER BY bucket ASC";
-
-            let mut stmt = conn.prepare(sql)?;
-            let rows = stmt.query_map([start_ts], |row| {
-                Ok(DailyStats {
-                    date: row.get(0)?,
-                    request_count: row.get::<_, i64>(1)? as u64,
-                    total_cost: format!("{:.6}", row.get::<_, f64>(2)?),
-                    total_tokens: row.get::<_, i64>(3)? as u64,
-                    total_input_tokens: row.get::<_, i64>(4)? as u64,
-                    total_output_tokens: row.get::<_, i64>(5)? as u64,
-                    total_cache_creation_tokens: row.get::<_, i64>(6)? as u64,
-                    total_cache_read_tokens: row.get::<_, i64>(7)? as u64,
-                })
-            })?;
-
-            let mut map = HashMap::new();
-            for row in rows {
-                let stat = row?;
-                map.insert(stat.date.clone(), stat);
-            }
-
-            let mut stats = Vec::new();
-
-            for i in 0..days {
-                let day = start_day + Duration::days(i as i64);
-                let key = day.format("%Y-%m-%dT00:00:00").to_string();
-                if let Some(stat) = map.remove(&key) {
-                    stats.push(stat);
-                } else {
-                    stats.push(DailyStats {
-                        date: key,
-                        request_count: 0,
-                        total_cost: "0.000000".to_string(),
-                        total_tokens: 0,
-                        total_input_tokens: 0,
-                        total_output_tokens: 0,
-                        total_cache_creation_tokens: 0,
-                        total_cache_read_tokens: 0,
-                    });
-                }
-            }
-            Ok(stats)
+        if start_ts >= end_ts {
+            start_ts = end_ts - 24 * 60 * 60;
         }
+
+        let duration = end_ts - start_ts;
+        let bucket_seconds: i64 = if duration <= 24 * 60 * 60 {
+            60 * 60
+        } else {
+            24 * 60 * 60
+        };
+        let mut bucket_count: i64 = if duration <= 0 {
+            1
+        } else {
+            ((duration as f64) / bucket_seconds as f64).ceil() as i64
+        };
+
+        // 固定 24 小时窗口为 24 个小时桶，避免浮点误差
+        if bucket_seconds == 60 * 60 {
+            bucket_count = 24;
+        }
+
+        if bucket_count < 1 {
+            bucket_count = 1;
+        }
+
+        let sql = "
+            SELECT 
+                CAST((created_at - ?1) / ?3 AS INTEGER) as bucket_idx,
+                COUNT(*) as request_count,
+                COALESCE(SUM(CAST(total_cost_usd AS REAL)), 0) as total_cost,
+                COALESCE(SUM(input_tokens + output_tokens), 0) as total_tokens,
+                COALESCE(SUM(input_tokens), 0) as total_input_tokens,
+                COALESCE(SUM(output_tokens), 0) as total_output_tokens,
+                COALESCE(SUM(cache_creation_tokens), 0) as total_cache_creation_tokens,
+                COALESCE(SUM(cache_read_tokens), 0) as total_cache_read_tokens
+            FROM proxy_request_logs
+            WHERE created_at >= ?1 AND created_at <= ?2
+            GROUP BY bucket_idx
+            ORDER BY bucket_idx ASC";
+
+        let mut stmt = conn.prepare(sql)?;
+        let rows = stmt.query_map(params![start_ts, end_ts, bucket_seconds], |row| {
+            Ok((
+                row.get::<_, i64>(0)?,
+                DailyStats {
+                    date: String::new(),
+                    request_count: row.get::<_, i64>(1)? as u64,
+                    total_cost: format!("{:.6}", row.get::<_, f64>(2)?),
+                    total_tokens: row.get::<_, i64>(3)? as u64,
+                    total_input_tokens: row.get::<_, i64>(4)? as u64,
+                    total_output_tokens: row.get::<_, i64>(5)? as u64,
+                    total_cache_creation_tokens: row.get::<_, i64>(6)? as u64,
+                    total_cache_read_tokens: row.get::<_, i64>(7)? as u64,
+                },
+            ))
+        })?;
+
+        let mut map: HashMap<i64, DailyStats> = HashMap::new();
+        for row in rows {
+            let (mut bucket_idx, stat) = row?;
+            if bucket_idx < 0 {
+                continue;
+            }
+            if bucket_idx >= bucket_count {
+                bucket_idx = bucket_count - 1;
+            }
+            map.insert(bucket_idx, stat);
+        }
+
+        let mut stats = Vec::with_capacity(bucket_count as usize);
+        for i in 0..bucket_count {
+            let bucket_start_ts = start_ts + i * bucket_seconds;
+            let bucket_start = Local
+                .timestamp_opt(bucket_start_ts, 0)
+                .single()
+                .unwrap_or_else(Local::now);
+
+            let date = bucket_start.format("%Y-%m-%dT%H:%M:%S").to_string();
+
+            if let Some(mut stat) = map.remove(&i) {
+                stat.date = date;
+                stats.push(stat);
+            } else {
+                stats.push(DailyStats {
+                    date,
+                    request_count: 0,
+                    total_cost: "0.000000".to_string(),
+                    total_tokens: 0,
+                    total_input_tokens: 0,
+                    total_output_tokens: 0,
+                    total_cache_creation_tokens: 0,
+                    total_cache_read_tokens: 0,
+                });
+            }
+        }
+
+        Ok(stats)
     }
 
     /// 获取 Provider 统计
@@ -829,89 +798,46 @@ impl Database {
     }
 }
 
-/// 标准化模型名称：去除供应商前缀并将点号替换为短横线
-/// 例如：anthropic/claude-haiku-4.5 → claude-haiku-4-5
-fn normalize_model_id(model_id: &str) -> String {
-    // 1. 去除供应商前缀（如 anthropic/、openai/）
-    let stripped = if let Some(pos) = model_id.find('/') {
-        &model_id[pos + 1..]
-    } else {
-        model_id
-    };
-    // 2. 将点号替换为短横线（如 claude-haiku-4.5 → claude-haiku-4-5）
-    stripped.replace('.', "-")
-}
-
 pub(crate) fn find_model_pricing_row(
     conn: &Connection,
     model_id: &str,
 ) -> Result<Option<(String, String, String, String)>, AppError> {
-    // 0. 标准化模型名称（去除前缀 + 点号转短横线）
-    // 例如：anthropic/claude-haiku-4.5 → claude-haiku-4-5
-    let normalized = normalize_model_id(model_id);
+    // 1) 去除供应商前缀（/ 之前）与冒号后缀（: 之后），例如 moonshotai/kimi-k2-0905:exa → kimi-k2-0905
+    let without_prefix = model_id
+        .rsplit_once('/')
+        .map(|(_, rest)| rest)
+        .unwrap_or(model_id);
+    let cleaned = without_prefix
+        .split(':')
+        .next()
+        .map(str::trim)
+        .unwrap_or(without_prefix);
 
-    // 1. 精确匹配（先尝试原始名称，再尝试标准化后的名称）
-    for id in [model_id, normalized.as_str()] {
-        let exact = conn
-            .query_row(
-                "SELECT input_cost_per_million, output_cost_per_million,
-                        cache_read_cost_per_million, cache_creation_cost_per_million
-                 FROM model_pricing
-                 WHERE model_id = ?1",
-                [id],
-                |row| {
-                    Ok((
-                        row.get::<_, String>(0)?,
-                        row.get::<_, String>(1)?,
-                        row.get::<_, String>(2)?,
-                        row.get::<_, String>(3)?,
-                    ))
-                },
-            )
-            .optional()
-            .map_err(|e| AppError::Database(format!("查询模型定价失败: {e}")))?;
+    // 2) 精确匹配清洗后的名称
+    let exact = conn
+        .query_row(
+            "SELECT input_cost_per_million, output_cost_per_million,
+                    cache_read_cost_per_million, cache_creation_cost_per_million
+             FROM model_pricing
+             WHERE model_id = ?1",
+            [cleaned],
+            |row| {
+                Ok((
+                    row.get::<_, String>(0)?,
+                    row.get::<_, String>(1)?,
+                    row.get::<_, String>(2)?,
+                    row.get::<_, String>(3)?,
+                ))
+            },
+        )
+        .optional()
+        .map_err(|e| AppError::Database(format!("查询模型定价失败: {e}")))?;
 
-        if exact.is_some() {
-            if id != model_id {
-                log::info!("模型 {model_id} 标准化后精确匹配到: {id}");
-            }
-            return Ok(exact);
-        }
+    if exact.is_none() {
+        log::warn!("模型 {model_id}（清洗后: {cleaned}）未找到定价信息，成本将记录为 0");
     }
 
-    // 2. 逐步删除后缀匹配（claude-haiku-4-5-20250929 → claude-haiku-4-5 → claude-haiku-4 → claude-haiku）
-    // 使用标准化后的名称进行后缀匹配
-    let mut current = normalized;
-    while let Some(pos) = current.rfind('-') {
-        current = current[..pos].to_string();
-
-        let result = conn
-            .query_row(
-                "SELECT input_cost_per_million, output_cost_per_million,
-                        cache_read_cost_per_million, cache_creation_cost_per_million
-                 FROM model_pricing
-                 WHERE model_id = ?1",
-                [&current],
-                |row| {
-                    Ok((
-                        row.get::<_, String>(0)?,
-                        row.get::<_, String>(1)?,
-                        row.get::<_, String>(2)?,
-                        row.get::<_, String>(3)?,
-                    ))
-                },
-            )
-            .optional()
-            .map_err(|e| AppError::Database(format!("查询模型定价失败: {e}")))?;
-
-        if result.is_some() {
-            log::info!("模型 {model_id} 通过删除后缀匹配到: {current}");
-            return Ok(result);
-        }
-    }
-
-    log::warn!("模型 {model_id} 未找到定价信息，成本将记录为 0");
-    Ok(None)
+    Ok(exact)
 }
 
 #[cfg(test)]
@@ -991,54 +917,39 @@ mod tests {
         let db = Database::memory()?;
         let conn = lock_conn!(db.conn);
 
-        // 测试精确匹配
-        let result = find_model_pricing_row(&conn, "claude-sonnet-4-5")?;
-        assert!(result.is_some(), "应该能精确匹配 claude-sonnet-4-5");
+        // 准备额外定价数据，覆盖前缀/后缀清洗场景
+        conn.execute(
+            "INSERT OR REPLACE INTO model_pricing (
+                model_id, display_name, input_cost_per_million, output_cost_per_million,
+                cache_read_cost_per_million, cache_creation_cost_per_million
+            ) VALUES (?, ?, ?, ?, ?, ?)",
+            params![
+                "claude-haiku-4.5",
+                "Claude Haiku 4.5",
+                "1.0",
+                "2.0",
+                "0.0",
+                "0.0"
+            ],
+        )?;
 
-        // 测试带供应商前缀的模型名称（anthropic/claude-haiku-4.5 → claude-haiku-4-5）
-        let result = find_model_pricing_row(&conn, "anthropic/claude-haiku-4.5")?;
-        assert!(
-            result.is_some(),
-            "应该能匹配带前缀的模型 anthropic/claude-haiku-4.5"
-        );
-
-        // 测试带供应商前缀 + 点号的模型名称
-        let result = find_model_pricing_row(&conn, "anthropic/claude-sonnet-4.5")?;
-        assert!(
-            result.is_some(),
-            "应该能匹配带前缀的模型 anthropic/claude-sonnet-4.5"
-        );
-
-        // 测试逐步删除后缀匹配 - 日期后缀
-        let result = find_model_pricing_row(&conn, "claude-sonnet-4-5-20241022")?;
-        assert!(
-            result.is_some(),
-            "应该能通过删除后缀匹配 claude-sonnet-4-5-20241022"
-        );
-
-        // 测试逐步删除后缀匹配 - 多个后缀
-        let result = find_model_pricing_row(&conn, "claude-haiku-4-5-20240229-preview")?;
-        assert!(
-            result.is_some(),
-            "应该能通过删除后缀匹配 claude-haiku-4-5-20240229-preview"
-        );
-
-        // 测试 GPT 模型
-        let result = find_model_pricing_row(&conn, "gpt-5-2024-11-20")?;
-        assert!(result.is_some(), "应该能通过删除后缀匹配 gpt-5-2024-11-20");
-
-        // 测试 Gemini 模型
-        let result = find_model_pricing_row(&conn, "gemini-2.5-flash-exp")?;
-        assert!(
-            result.is_some(),
-            "应该能通过删除后缀匹配 gemini-2.5-flash-exp"
-        );
-
-        // 测试 claude-sonnet-4-5 命名格式
+        // 测试精确匹配（seed_model_pricing 已预置 claude-sonnet-4-5-20250929）
         let result = find_model_pricing_row(&conn, "claude-sonnet-4-5-20250929")?;
         assert!(
             result.is_some(),
-            "应该能通过删除后缀匹配 claude-sonnet-4-5-20250929"
+            "应该能精确匹配 claude-sonnet-4-5-20250929"
+        );
+
+        // 清洗：去除前缀和冒号后缀
+        let result = find_model_pricing_row(&conn, "anthropic/claude-haiku-4.5")?;
+        assert!(
+            result.is_some(),
+            "带前缀的模型 anthropic/claude-haiku-4.5 应能匹配到 claude-haiku-4.5"
+        );
+        let result = find_model_pricing_row(&conn, "moonshotai/kimi-k2-0905:exa")?;
+        assert!(
+            result.is_some(),
+            "带前缀+冒号后缀的模型应清洗后匹配到 kimi-k2-0905"
         );
 
         // 测试不存在的模型