fix(billing): 修复 OpenAI fast 档位计费并补齐展示
- 打通 service_tier 在 OpenAI HTTP、WS、passthrough 与 usage 记录中的传递 - 修正 priority/flex 计费逻辑,并将 fast 归一化为 priority - 在用户端和管理端补齐服务档位与计费明细展示 - 补齐前后端测试,并修复 WS 限流信号重复持久化导致的全量回归失败 Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
This commit is contained in:
@@ -1,11 +1,40 @@
|
||||
package service
|
||||
|
||||
import (
|
||||
"encoding/json"
|
||||
"testing"
|
||||
|
||||
"github.com/stretchr/testify/require"
|
||||
)
|
||||
|
||||
func TestParsePricingData_ParsesPriorityAndServiceTierFields(t *testing.T) {
|
||||
svc := &PricingService{}
|
||||
body := []byte(`{
|
||||
"gpt-5.4": {
|
||||
"input_cost_per_token": 0.0000025,
|
||||
"input_cost_per_token_priority": 0.000005,
|
||||
"output_cost_per_token": 0.000015,
|
||||
"output_cost_per_token_priority": 0.00003,
|
||||
"cache_creation_input_token_cost": 0.0000025,
|
||||
"cache_read_input_token_cost": 0.00000025,
|
||||
"cache_read_input_token_cost_priority": 0.0000005,
|
||||
"supports_service_tier": true,
|
||||
"supports_prompt_caching": true,
|
||||
"litellm_provider": "openai",
|
||||
"mode": "chat"
|
||||
}
|
||||
}`)
|
||||
|
||||
data, err := svc.parsePricingData(body)
|
||||
require.NoError(t, err)
|
||||
pricing := data["gpt-5.4"]
|
||||
require.NotNil(t, pricing)
|
||||
require.InDelta(t, 5e-6, pricing.InputCostPerTokenPriority, 1e-12)
|
||||
require.InDelta(t, 3e-5, pricing.OutputCostPerTokenPriority, 1e-12)
|
||||
require.InDelta(t, 5e-7, pricing.CacheReadInputTokenCostPriority, 1e-12)
|
||||
require.True(t, pricing.SupportsServiceTier)
|
||||
}
|
||||
|
||||
func TestGetModelPricing_Gpt53CodexSparkUsesGpt51CodexPricing(t *testing.T) {
|
||||
sparkPricing := &LiteLLMModelPricing{InputCostPerToken: 1}
|
||||
gpt53Pricing := &LiteLLMModelPricing{InputCostPerToken: 9}
|
||||
@@ -68,3 +97,64 @@ func TestGetModelPricing_Gpt54UsesStaticFallbackWhenRemoteMissing(t *testing.T)
|
||||
require.InDelta(t, 2.0, got.LongContextInputCostMultiplier, 1e-12)
|
||||
require.InDelta(t, 1.5, got.LongContextOutputCostMultiplier, 1e-12)
|
||||
}
|
||||
|
||||
func TestParsePricingData_PreservesPriorityAndServiceTierFields(t *testing.T) {
|
||||
raw := map[string]any{
|
||||
"gpt-5.4": map[string]any{
|
||||
"input_cost_per_token": 2.5e-6,
|
||||
"input_cost_per_token_priority": 5e-6,
|
||||
"output_cost_per_token": 15e-6,
|
||||
"output_cost_per_token_priority": 30e-6,
|
||||
"cache_read_input_token_cost": 0.25e-6,
|
||||
"cache_read_input_token_cost_priority": 0.5e-6,
|
||||
"supports_service_tier": true,
|
||||
"supports_prompt_caching": true,
|
||||
"litellm_provider": "openai",
|
||||
"mode": "chat",
|
||||
},
|
||||
}
|
||||
body, err := json.Marshal(raw)
|
||||
require.NoError(t, err)
|
||||
|
||||
svc := &PricingService{}
|
||||
pricingMap, err := svc.parsePricingData(body)
|
||||
require.NoError(t, err)
|
||||
|
||||
pricing := pricingMap["gpt-5.4"]
|
||||
require.NotNil(t, pricing)
|
||||
require.InDelta(t, 2.5e-6, pricing.InputCostPerToken, 1e-12)
|
||||
require.InDelta(t, 5e-6, pricing.InputCostPerTokenPriority, 1e-12)
|
||||
require.InDelta(t, 15e-6, pricing.OutputCostPerToken, 1e-12)
|
||||
require.InDelta(t, 30e-6, pricing.OutputCostPerTokenPriority, 1e-12)
|
||||
require.InDelta(t, 0.25e-6, pricing.CacheReadInputTokenCost, 1e-12)
|
||||
require.InDelta(t, 0.5e-6, pricing.CacheReadInputTokenCostPriority, 1e-12)
|
||||
require.True(t, pricing.SupportsServiceTier)
|
||||
}
|
||||
|
||||
func TestParsePricingData_PreservesServiceTierPriorityFields(t *testing.T) {
|
||||
svc := &PricingService{}
|
||||
pricingData, err := svc.parsePricingData([]byte(`{
|
||||
"gpt-5.4": {
|
||||
"input_cost_per_token": 0.0000025,
|
||||
"input_cost_per_token_priority": 0.000005,
|
||||
"output_cost_per_token": 0.000015,
|
||||
"output_cost_per_token_priority": 0.00003,
|
||||
"cache_read_input_token_cost": 0.00000025,
|
||||
"cache_read_input_token_cost_priority": 0.0000005,
|
||||
"supports_service_tier": true,
|
||||
"litellm_provider": "openai",
|
||||
"mode": "chat"
|
||||
}
|
||||
}`))
|
||||
require.NoError(t, err)
|
||||
|
||||
pricing := pricingData["gpt-5.4"]
|
||||
require.NotNil(t, pricing)
|
||||
require.InDelta(t, 0.0000025, pricing.InputCostPerToken, 1e-12)
|
||||
require.InDelta(t, 0.000005, pricing.InputCostPerTokenPriority, 1e-12)
|
||||
require.InDelta(t, 0.000015, pricing.OutputCostPerToken, 1e-12)
|
||||
require.InDelta(t, 0.00003, pricing.OutputCostPerTokenPriority, 1e-12)
|
||||
require.InDelta(t, 0.00000025, pricing.CacheReadInputTokenCost, 1e-12)
|
||||
require.InDelta(t, 0.0000005, pricing.CacheReadInputTokenCostPriority, 1e-12)
|
||||
require.True(t, pricing.SupportsServiceTier)
|
||||
}
|
||||
|
||||
Reference in New Issue
Block a user