Jaymr/new-api
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

feat: bill OpenAI cache_write_tokens at cache-creation price with zero clamp

Browse Source 
Parse OpenAI's native cache_write_tokens (chat prompt_tokens_details /
responses input_tokens_details), bill it at the cache-creation ratio, and
clamp the uncached prompt remainder at zero since cached + cache-write can
exceed prompt_tokens. Propagate the field through chat/responses/claude
format conversions and tiered expression billing (cc variable).
This commit is contained in:
CaIon
2026-07-11 21:18:49 +08:00
parent c36418c863
commit 48068ce923
18 changed files with 158 additions and 20 deletions
Download Patch File Download Diff File Expand all files Collapse all files
dto/billing_usage.go
+1
View File
@@ -88,6 +88,7 @@ func HasOpenAIUsageTokens(usage *Usage) bool {
}
if usage.PromptTokensDetails.CachedTokens != 0 ||
usage.PromptTokensDetails.CachedCreationTokens != 0 ||
usage.PromptTokensDetails.CacheWriteTokens != 0 ||
usage.PromptTokensDetails.TextTokens != 0 ||
usage.PromptTokensDetails.ImageTokens != 0 ||
usage.PromptTokensDetails.AudioTokens != 0 {
dto/openai_request.go
+2
View File
@@ -849,6 +849,7 @@ type OpenAIResponsesRequest struct {
MaxOutputTokens *uint `json:"max_output_tokens,omitempty"`
TopLogProbs *int `json:"top_logprobs,omitempty"`
Metadata json.RawMessage `json:"metadata,omitempty"`
Moderation json.RawMessage `json:"moderation,omitempty"`
ParallelToolCalls json.RawMessage `json:"parallel_tool_calls,omitempty"`
PreviousResponseID string `json:"previous_response_id,omitempty"`
Reasoning *Reasoning `json:"reasoning,omitempty"`
@@ -859,6 +860,7 @@ type OpenAIResponsesRequest struct {
// This field is allowed by default and can be disabled via channel setting disable_store.
Store json.RawMessage `json:"store,omitempty"`
PromptCacheKey json.RawMessage `json:"prompt_cache_key,omitempty"`
PromptCacheOptions json.RawMessage `json:"prompt_cache_options,omitempty"`
PromptCacheRetention json.RawMessage `json:"prompt_cache_retention,omitempty"`
// SafetyIdentifier carries client identity for policy abuse detection.
// This field is filtered by default and can be enabled via channel setting allow_safety_identifier.
dto/openai_response.go
+20 -3
View File
@@ -256,9 +256,26 @@ type OpenAIVideoResponse struct {
type InputTokenDetails struct {
CachedTokens int `json:"cached_tokens"`
CachedCreationTokens int `json:"cached_creation_tokens,omitempty"`
TextTokens int `json:"text_tokens"`
AudioTokens int `json:"audio_tokens"`
ImageTokens int `json:"image_tokens"`
// CacheWriteTokens is OpenAI's native cache-write count, reported as
// prompt_tokens_details.cache_write_tokens (Chat Completions) or
// input_tokens_details.cache_write_tokens (Responses). It is billed at the
// cache-creation price.
CacheWriteTokens int `json:"cache_write_tokens,omitempty"`
TextTokens int `json:"text_tokens"`
AudioTokens int `json:"audio_tokens"`
ImageTokens int `json:"image_tokens"`
}
// CacheCreationTokensTotal returns the cache-write token count regardless of
// which field the upstream reported it in: Claude-derived conversions populate
// CachedCreationTokens while OpenAI reports cache_write_tokens natively. Both
// are billed at the cache-creation price; when both are present the larger
// value wins so the same tokens are never double-counted.
func (d InputTokenDetails) CacheCreationTokensTotal() int {
if d.CacheWriteTokens > d.CachedCreationTokens {
return d.CacheWriteTokens
}
return d.CachedCreationTokens
}
type OutputTokenDetails struct {
dto/openai_responses_compaction_request.go
+8 -6
View File
@@ -17,12 +17,14 @@ type OpenAIResponsesCompactionRequest struct {
// Codex compact request parity:
// https://github.com/openai/codex/commit/53d59722268dde82fb93c1f37964ce196c2a86d7
// https://github.com/openai/codex/commit/5d6f23a27bf9c90709af527a7108c1c2eadf5123
Tools json.RawMessage `json:"tools,omitempty"`
ParallelToolCalls json.RawMessage `json:"parallel_tool_calls,omitempty"`
Reasoning *Reasoning `json:"reasoning,omitempty"`
ServiceTier string `json:"service_tier,omitempty"`
PromptCacheKey json.RawMessage `json:"prompt_cache_key,omitempty"`
Text json.RawMessage `json:"text,omitempty"`
Tools json.RawMessage `json:"tools,omitempty"`
ParallelToolCalls json.RawMessage `json:"parallel_tool_calls,omitempty"`
Reasoning *Reasoning `json:"reasoning,omitempty"`
ServiceTier string `json:"service_tier,omitempty"`
PromptCacheKey json.RawMessage `json:"prompt_cache_key,omitempty"`
PromptCacheOptions json.RawMessage `json:"prompt_cache_options,omitempty"`
PromptCacheRetention json.RawMessage `json:"prompt_cache_retention,omitempty"`
Text json.RawMessage `json:"text,omitempty"`
}
func (r *OpenAIResponsesCompactionRequest) GetTokenCountMeta() *types.TokenCountMeta {