diff --git a/apps/api/src/services/ai_rate_limiter.py b/apps/api/src/services/ai_rate_limiter.py index e1267817..14922841 100644 --- a/apps/api/src/services/ai_rate_limiter.py +++ b/apps/api/src/services/ai_rate_limiter.py @@ -8,10 +8,12 @@ AI Rate Limiter - Gemini API 用量閥值控制 - 每分鐘請求限制 (RPM) - 每日請求限制 - 每日 Token 限制 -- 超限自動降級 +- 🔴 累積成本限制 ($5 USD) - 2026-03-29 ogt 新增 +- 超限自動降級 + Telegram 告警 -版本: v1.0 +版本: v1.1 建立日期: 2026-03-26 21:00 (台北時區) +更新日期: 2026-03-29 22:45 (台北時區) 建立者: Claude Code """ @@ -37,11 +39,35 @@ RATE_LIMITS = { }, } +# ============================================================================= +# 2026-03-29 ogt: 累積成本限制 (統帥要求) +# ============================================================================= + +COST_LIMITS = { + "gemini": { + "total_cost_usd": 5.0, # 🔴 總成本上限 $5 USD，超過自動停用 + "alert_threshold_usd": 4.0, # 警告閾值 $4 USD + }, + "claude": { + "total_cost_usd": 10.0, + "alert_threshold_usd": 8.0, + }, +} + +# Gemini 1.5 Flash 定價 (per token) +GEMINI_PRICING = { + "input_per_token": 0.000000075, # $0.075 / 1M tokens + "output_per_token": 0.0000003, # $0.30 / 1M tokens +} + # Redis Keys REDIS_KEY_PREFIX = "ai_rate:" RPM_KEY = f"{REDIS_KEY_PREFIX}rpm:{{provider}}" DAILY_REQ_KEY = f"{REDIS_KEY_PREFIX}daily_req:{{provider}}:{{date}}" DAILY_TOKEN_KEY = f"{REDIS_KEY_PREFIX}daily_token:{{provider}}:{{date}}" +# 2026-03-29 ogt: 累積成本 Key (不過期，手動重置) +TOTAL_COST_KEY = f"{REDIS_KEY_PREFIX}total_cost:{{provider}}" +COST_ALERT_SENT_KEY = f"{REDIS_KEY_PREFIX}cost_alert_sent:{{provider}}" # ============================================================================= @@ -100,6 +126,25 @@ class AIRateLimiter: r = await self._get_redis() today = self._get_today() + # 0. 🔴 2026-03-29 ogt: 檢查累積成本 (最高優先級) + if provider in COST_LIMITS: + cost_limit = COST_LIMITS[provider]["total_cost_usd"] + total_cost_key = TOTAL_COST_KEY.format(provider=provider) + current_cost = await r.get(total_cost_key) + current_cost = float(current_cost) if current_cost else 0.0 + + if current_cost >= cost_limit: + logger.error( + "ai_cost_limit_exceeded_blocking", + provider=provider, + current_cost=f"${current_cost:.4f}", + limit=f"${cost_limit:.2f}", + action="AUTO_SWITCH_TO_OLLAMA", + ) + # 發送告警 (只發一次) + await self._send_cost_alert(provider, current_cost, cost_limit) + return False, f"🔴 成本超限! ${current_cost:.2f} >= ${cost_limit:.2f}，已自動切換到 Ollama" + # 1. 檢查 RPM rpm_key = RPM_KEY.format(provider=provider) current_rpm = await r.get(rpm_key) @@ -169,6 +214,140 @@ class AIRateLimiter: return True, None + async def record_cost(self, provider: str, cost_usd: float) -> None: + """ + 2026-03-29 ogt: 記錄累積成本 + + Args: + provider: AI 提供者 + cost_usd: 本次成本 (USD) + """ + if provider not in COST_LIMITS or cost_usd <= 0: + return + + r = await self._get_redis() + total_cost_key = TOTAL_COST_KEY.format(provider=provider) + + # 使用 INCRBYFLOAT 原子操作 + new_total = await r.incrbyfloat(total_cost_key, cost_usd) + + logger.info( + "ai_cost_recorded", + provider=provider, + cost_usd=f"${cost_usd:.6f}", + total_cost=f"${new_total:.4f}", + ) + + # 檢查是否需要發送警告 (接近上限) + alert_threshold = COST_LIMITS[provider]["alert_threshold_usd"] + if new_total >= alert_threshold: + await self._send_cost_warning(provider, new_total, alert_threshold) + + async def _send_cost_alert(self, provider: str, current_cost: float, limit: float) -> None: + """ + 2026-03-29 ogt: 發送成本超限告警到 Telegram (只發一次) + """ + r = await self._get_redis() + alert_sent_key = COST_ALERT_SENT_KEY.format(provider=provider) + + # 檢查是否已發送 + if await r.get(alert_sent_key): + return + + # 標記已發送 (24小時後可重新發送) + await r.set(alert_sent_key, "1", ex=86400) + + try: + from src.core.config import settings + + if not settings.OPENCLAW_TG_BOT_TOKEN or not settings.OPENCLAW_TG_CHAT_ID: + logger.warning("telegram_not_configured_for_cost_alert") + return + + import httpx + + message = ( + f"🚨🚨🚨 AI 成本超限警報 🚨🚨🚨\n\n" + f"Provider: {provider.upper()}\n" + f"累積成本: ${current_cost:.2f}\n" + f"上限: ${limit:.2f}\n\n" + f"⚡ 已自動切換到 Ollama\n\n" + f"如需恢復 {provider.upper()}，請執行:\n" + f"redis-cli DEL ai_rate:total_cost:{provider}" + ) + + async with httpx.AsyncClient(timeout=10.0) as client: + await client.post( + f"https://api.telegram.org/bot{settings.OPENCLAW_TG_BOT_TOKEN}/sendMessage", + json={ + "chat_id": settings.OPENCLAW_TG_CHAT_ID, + "text": message, + "parse_mode": "HTML", + }, + ) + + logger.error( + "ai_cost_alert_sent", + provider=provider, + current_cost=f"${current_cost:.2f}", + limit=f"${limit:.2f}", + ) + + except Exception as e: + logger.error("ai_cost_alert_failed", error=str(e)) + + async def _send_cost_warning(self, provider: str, current_cost: float, threshold: float) -> None: + """ + 2026-03-29 ogt: 發送成本接近上限警告 + """ + r = await self._get_redis() + warning_key = f"{REDIS_KEY_PREFIX}cost_warning_sent:{provider}" + + # 每小時只發一次警告 + if await r.get(warning_key): + return + + await r.set(warning_key, "1", ex=3600) + + try: + from src.core.config import settings + + if not settings.OPENCLAW_TG_BOT_TOKEN or not settings.OPENCLAW_TG_CHAT_ID: + return + + import httpx + + limit = COST_LIMITS[provider]["total_cost_usd"] + remaining = limit - current_cost + + message = ( + f"⚠️ AI 成本警告\n\n" + f"Provider: {provider.upper()}\n" + f"累積成本: ${current_cost:.2f} / ${limit:.2f}\n" + f"剩餘額度: ${remaining:.2f}\n\n" + f"接近上限，請注意監控！" + ) + + async with httpx.AsyncClient(timeout=10.0) as client: + await client.post( + f"https://api.telegram.org/bot{settings.OPENCLAW_TG_BOT_TOKEN}/sendMessage", + json={ + "chat_id": settings.OPENCLAW_TG_CHAT_ID, + "text": message, + "parse_mode": "HTML", + }, + ) + + logger.warning( + "ai_cost_warning_sent", + provider=provider, + current_cost=f"${current_cost:.2f}", + threshold=f"${threshold:.2f}", + ) + + except Exception as e: + logger.warning("ai_cost_warning_failed", error=str(e)) + async def record_tokens(self, provider: str, tokens: int) -> None: """ 記錄 Token 用量 (回應後呼叫) @@ -195,7 +374,7 @@ class AIRateLimiter: async def get_usage_stats(self, provider: str) -> dict: """ - 取得用量統計 + 取得用量統計 (含成本) Args: provider: AI 提供者 @@ -213,10 +392,27 @@ class AIRateLimiter: rpm_key = RPM_KEY.format(provider=provider) daily_req_key = DAILY_REQ_KEY.format(provider=provider, date=today) daily_token_key = DAILY_TOKEN_KEY.format(provider=provider, date=today) + total_cost_key = TOTAL_COST_KEY.format(provider=provider) current_rpm = await r.get(rpm_key) current_daily = await r.get(daily_req_key) current_tokens = await r.get(daily_token_key) + current_cost = await r.get(total_cost_key) + + # 2026-03-29 ogt: 加入成本資訊 + cost_info = {} + if provider in COST_LIMITS: + cost_limit = COST_LIMITS[provider] + current_cost_float = float(current_cost) if current_cost else 0.0 + cost_info = { + "total_cost_usd": { + "current": round(current_cost_float, 4), + "limit": cost_limit["total_cost_usd"], + "remaining": round(cost_limit["total_cost_usd"] - current_cost_float, 4), + "alert_threshold": cost_limit["alert_threshold_usd"], + }, + "cost_exceeded": current_cost_float >= cost_limit["total_cost_usd"], + } return { "provider": provider, @@ -233,8 +429,23 @@ class AIRateLimiter: "current": int(current_tokens) if current_tokens else 0, "limit": limits["daily_tokens"], }, + **cost_info, } + async def reset_cost(self, provider: str) -> None: + """ + 2026-03-29 ogt: 重置累積成本 (統帥授權後使用) + + Args: + provider: AI 提供者 + """ + r = await self._get_redis() + total_cost_key = TOTAL_COST_KEY.format(provider=provider) + alert_sent_key = COST_ALERT_SENT_KEY.format(provider=provider) + + await r.delete(total_cost_key, alert_sent_key) + logger.info("ai_cost_reset", provider=provider) + async def reset_limits(self, provider: str) -> None: """ 重置限制 (緊急用) diff --git a/apps/api/src/services/openclaw.py b/apps/api/src/services/openclaw.py index 3a03b101..d3ef3afa 100644 --- a/apps/api/src/services/openclaw.py +++ b/apps/api/src/services/openclaw.py @@ -908,6 +908,11 @@ class OpenClawService: ) # Langfuse: 記錄成功評分 trace.score(name="provider_success", value=1.0, comment=f"Success via {provider}") + + # 2026-03-29 ogt: 記錄累積成本 (Gemini/Claude) + if cost_usd > 0: + await rate_limiter.record_cost(provider, cost_usd) + return response, provider, True, total_tokens, cost_usd logger.warning("ai_provider_failed_fallback", provider=provider, latency_ms=latency_ms)