diff --git a/common/model-ratio.go b/common/model-ratio.go
index 08cde8c70..b1f6df4b2 100644
--- a/common/model-ratio.go
+++ b/common/model-ratio.go
@@ -78,8 +78,8 @@ var ModelRatio = map[string]float64{
 	"text-embedding-3-small":    0.01,
 	"text-embedding-3-large":    0.065,
 	"text-search-ada-doc-001":   10,
-	"text-moderation-stable":    0.1,
-	"text-moderation-latest":    0.1,
+	"text-moderation-stable":    0,		 // free to use
+	"text-moderation-latest":    0,	     // free to use
 	"dall-e-2":                  8,      // $0.016 - $0.020 / image
 	"dall-e-3":                  20,     // $0.040 - $0.120 / image
 	"claude-instant-1":          0.815,  // $1.63 / 1M tokens
diff --git a/relay/controller/helper.go b/relay/controller/helper.go
index 6154f291a..1b02cd40d 100644
--- a/relay/controller/helper.go
+++ b/relay/controller/helper.go
@@ -113,12 +113,18 @@ func postConsumeQuota(ctx context.Context, usage *openai.Usage, meta *util.Relay
 	if err != nil {
 		logger.Error(ctx, "error update user quota cache: "+err.Error())
 	}
-	if quota != 0 {
+
+	// if model ratio is zero, quota can be zero too
+	// without setting custom ratios, only "text-moderation-*" from OpenAI is currently 0
+	if quota >= 0 {
 		logContent := fmt.Sprintf("模型倍率 %.2f，分组倍率 %.2f，补全倍率 %.2f", modelRatio, groupRatio, completionRatio)
 		model.RecordConsumeLog(ctx, meta.UserId, meta.ChannelId, promptTokens, completionTokens, textRequest.Model, meta.TokenName, quota, logContent)
 		model.UpdateUserUsedQuotaAndRequestCount(meta.UserId, quota)
 		model.UpdateChannelUsedQuota(meta.ChannelId, quota)
 	}
+	if quota < 0 {
+		logger.Error(ctx, fmt.Sprintf("quota consumed is %d, something is wrong", quota))
+	}
 }
 
 func doRequest(ctx context.Context, c *gin.Context, meta *util.RelayMeta, isStream bool, fullRequestURL string, requestBody io.Reader) (*http.Response, error) {