ext/new-api
2
0
Fork 0
mirror of https://github.com/QuantumNous/new-api.git synced 2026-09-11 14:41:21 +00:00
Code Activity

fix: bound uncached remainder by prompt-max(cached,write) and forward compact prompt_cache_key

Browse Source
This commit is contained in:
CaIon
2026-07-11 22:18:26 +08:00
parent 48068ce923
commit 92d3c9d18f
9 changed files with 34 additions and 18 deletions
Download Patch File Download Diff File Expand all files Collapse all files
service/relayconvert/internal/oai_chat/to_claude_messages_resp.go
+2 -1
View File
@@ -44,7 +44,8 @@ func buildClaudeUsageFromOpenAIUsage(oaiUsage *dto.Usage) *dto.ClaudeUsage {
if oaiUsage.PromptTokensDetails.CacheWriteTokens > 0 {
// OpenAI native cache-write usage counts cached and cache-write tokens
// inside prompt_tokens, while Claude semantics reports input_tokens
// excluding both; the uncached remainder clamps at zero.
// excluding both. Both counts are unadjusted prefixes and may overlap,
// so clamp a negative remainder at zero.
inputTokens = oaiUsage.PromptTokens - oaiUsage.PromptTokensDetails.CachedTokens - cacheCreationTokens
if inputTokens < 0 {
inputTokens = 0
service/relayconvert/internal/oai_chat/to_claude_messages_resp_test.go
+1 -1
View File
@@ -92,7 +92,7 @@ func TestBuildClaudeUsageFromOpenAICacheWriteUsage(t *testing.T) {
require.NotNil(t, usage)
// Claude semantics reports input_tokens excluding cache read/write; the
// remainder 3619-2921-3616 clamps to 0.
// overlapping unadjusted prefixes drive the remainder negative, clamp to 0.
assert.Equal(t, 0, usage.InputTokens)
assert.Equal(t, 2921, usage.CacheReadInputTokens)
assert.Equal(t, 3616, usage.CacheCreationInputTokens)