50 lines
1.7 KiB
Go
50 lines
1.7 KiB
Go
package v1
|
|
|
|
import "github.com/gogf/gf/v2/frame/g"
|
|
|
|
// ModelComparisonReq compares 2-4 models across cost, latency, and success rate.
|
|
type ModelComparisonReq struct {
|
|
g.Meta `path:"/model-comparison" method:"get" tags:"TenantService" summary:"模型成本对比"`
|
|
Models string `json:"models" in:"query" v:"required" dc:"对比的模型名称，英文逗号分隔，2-4 个"`
|
|
Days int `json:"days" in:"query" v:"between:1,90" dc:"统计天数，默认 7" d:"7"`
|
|
}
|
|
|
|
type ModelComparisonRes struct {
|
|
Summary ModelComparisonSummary `json:"summary"`
|
|
Items []ModelComparisonItem `json:"items"`
|
|
Trends []ModelTrendDay `json:"trends"`
|
|
}
|
|
|
|
type ModelComparisonSummary struct {
|
|
TotalRequests int64 `json:"total_requests"`
|
|
TotalCost float64 `json:"total_cost"`
|
|
Recommended string `json:"recommended"`
|
|
Reason string `json:"reason"`
|
|
}
|
|
|
|
type ModelComparisonItem struct {
|
|
ModelName string `json:"model_name"`
|
|
Requests int64 `json:"requests"`
|
|
SuccessRate float64 `json:"success_rate"`
|
|
AvgLatencyMs float64 `json:"avg_latency_ms"`
|
|
P95LatencyMs float64 `json:"p95_latency_ms"`
|
|
TotalCost float64 `json:"total_cost"`
|
|
AvgCostPerRequest float64 `json:"avg_cost_per_request"`
|
|
InputTokens int64 `json:"input_tokens"`
|
|
OutputTokens int64 `json:"output_tokens"`
|
|
Score float64 `json:"score"`
|
|
IsRecommended bool `json:"is_recommended"`
|
|
}
|
|
|
|
type ModelTrendDay struct {
|
|
Date string `json:"date"`
|
|
Details []ModelTrendDayItem `json:"details"`
|
|
}
|
|
|
|
type ModelTrendDayItem struct {
|
|
ModelName string `json:"model_name"`
|
|
Requests int64 `json:"requests"`
|
|
Cost float64 `json:"cost"`
|
|
LatencyMs float64 `json:"latency_ms"`
|
|
}
|