github/claude-relay-service
1
0
Fork 0
mirror of https://github.com/Wei-Shaw/claude-relay-service.git synced 2026-01-23 17:39:16 +00:00
Code Issues Packages Projects Releases Wiki Activity

feat: openai转发增加apikey速率限制

Browse Source
This commit is contained in:
shaw
2025-10-08 08:36:43 +08:00
parent 9209f10fc3
commit eb304c7e70
5 changed files with 308 additions and 135 deletions
Download Patch File Download Diff File Expand all files Collapse all files

58
src/routes/openaiClaudeRoutes.js
View File

@@ -15,6 +15,7 @@ const apiKeyService = require('../services/apiKeyService')
const unifiedClaudeScheduler = require('../services/unifiedClaudeScheduler')
const claudeCodeHeadersService = require('../services/claudeCodeHeadersService')
const sessionHelper = require('../utils/sessionHelper')
const { updateRateLimitCounters } = require('../utils/rateLimitHelper')
// 加载模型定价数据
let modelPricingData = {}
@@ -33,6 +34,27 @@ function checkPermissions(apiKeyData, requiredPermission = 'claude') {
return permissions === 'all' || permissions === requiredPermission
}
function queueRateLimitUpdate(rateLimitInfo, usageSummary, model, context = '') {
if (!rateLimitInfo) {
return
}
const label = context ? ` (${context})` : ''
updateRateLimitCounters(rateLimitInfo, usageSummary, model)
.then(({ totalTokens, totalCost }) => {
if (totalTokens > 0) {
logger.api(`📊 Updated rate limit token count${label}: +${totalTokens} tokens`)
}
if (typeof totalCost === 'number' && totalCost > 0) {
logger.api(`💰 Updated rate limit cost count${label}: +$${totalCost.toFixed(6)}`)
}
})
.catch((error) => {
logger.error(`❌ Failed to update rate limit counters${label}:`, error)
})
}
// 📋 OpenAI 兼容的模型列表端点
router.get('/v1/models', authenticateApiKey, async (req, res) => {
try {
@@ -263,6 +285,12 @@ async function handleChatCompletion(req, res, apiKeyData) {
// 记录使用统计
if (usage && usage.input_tokens !== undefined && usage.output_tokens !== undefined) {
const model = usage.model || claudeRequest.model
const cacheCreateTokens =
(usage.cache_creation && typeof usage.cache_creation === 'object'
? (usage.cache_creation.ephemeral_5m_input_tokens || 0) +
(usage.cache_creation.ephemeral_1h_input_tokens || 0)
: usage.cache_creation_input_tokens || 0) || 0
const cacheReadTokens = usage.cache_read_input_tokens || 0
// 使用新的 recordUsageWithDetails 方法来支持详细的缓存数据
apiKeyService
@@ -275,6 +303,18 @@ async function handleChatCompletion(req, res, apiKeyData) {
.catch((error) => {
logger.error('❌ Failed to record usage:', error)
})
queueRateLimitUpdate(
req.rateLimitInfo,
{
inputTokens: usage.input_tokens || 0,
outputTokens: usage.output_tokens || 0,
cacheCreateTokens,
cacheReadTokens
},
model,
'openai-claude-stream'
)
}
},
// 流转换器
@@ -334,6 +374,12 @@ async function handleChatCompletion(req, res, apiKeyData) {
// 记录使用统计
if (claudeData.usage) {
const { usage } = claudeData
const cacheCreateTokens =
(usage.cache_creation && typeof usage.cache_creation === 'object'
? (usage.cache_creation.ephemeral_5m_input_tokens || 0) +
(usage.cache_creation.ephemeral_1h_input_tokens || 0)
: usage.cache_creation_input_tokens || 0) || 0
const cacheReadTokens = usage.cache_read_input_tokens || 0
// 使用新的 recordUsageWithDetails 方法来支持详细的缓存数据
apiKeyService
.recordUsageWithDetails(
@@ -345,6 +391,18 @@ async function handleChatCompletion(req, res, apiKeyData) {
.catch((error) => {
logger.error('❌ Failed to record usage:', error)
})
queueRateLimitUpdate(
req.rateLimitInfo,
{
inputTokens: usage.input_tokens || 0,
outputTokens: usage.output_tokens || 0,
cacheCreateTokens,
cacheReadTokens
},
claudeRequest.model,
'openai-claude-non-stream'
)
}
// 返回 OpenAI 格式响应