Skip to content

anthropic cache creation token mapping #1397

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Draft
wants to merge 2 commits into
base: main
Choose a base branch
from
Draft

anthropic cache creation token mapping #1397

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
4cf628c
fix anthropic cache creation token mapping in vertex ai
narengogi Oct 28, 2025
973ec2c
add null check
narengogi Oct 28, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
44 changes: 36 additions & 8 deletions src/providers/google-vertex-ai/chatComplete.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -772,7 +772,15 @@ export const VertexAnthropicChatCompleteResponseTransform: (
}

if ('content' in response) {
const { input_tokens = 0, output_tokens = 0 } = response?.usage ?? {};
const {
input_tokens = 0,
output_tokens = 0,
cache_creation_input_tokens,
cache_read_input_tokens,
} = response?.usage ?? {};

const shouldSendCacheUsage =
cache_creation_input_tokens || cache_read_input_tokens;
Copy link
Collaborator

@roh26it roh26it Nov 4, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

If both are zero, this would be false. Is that intended?


let content: AnthropicContentItem[] | string = strictOpenAiCompliance
? ''
Expand Down Expand Up @@ -827,7 +835,15 @@ export const VertexAnthropicChatCompleteResponseTransform: (
usage: {
prompt_tokens: input_tokens,
completion_tokens: output_tokens,
total_tokens: input_tokens + output_tokens,
total_tokens:
input_tokens +
output_tokens +
(cache_creation_input_tokens ?? 0) +
(cache_read_input_tokens ?? 0),
...(shouldSendCacheUsage && {
cache_read_input_tokens: cache_read_input_tokens,
cache_creation_input_tokens: cache_creation_input_tokens,
}),
},
};
}
Expand Down Expand Up @@ -895,11 +911,20 @@ export const VertexAnthropicChatCompleteStreamChunkTransform: (
);
}

const shouldSendCacheUsage =
parsedChunk.message?.usage?.cache_read_input_tokens ||
parsedChunk.message?.usage?.cache_creation_input_tokens;

if (parsedChunk.type === 'message_start' && parsedChunk.message?.usage) {
streamState.model = parsedChunk?.message?.model ?? '';

streamState.usage = {
prompt_tokens: parsedChunk.message.usage?.input_tokens,
prompt_tokens: parsedChunk.message?.usage?.input_tokens,
...(shouldSendCacheUsage && {
cache_read_input_tokens:
parsedChunk.message?.usage?.cache_read_input_tokens,
cache_creation_input_tokens:
parsedChunk.message?.usage?.cache_creation_input_tokens,
}),
};
return (
`data: ${JSON.stringify({
Expand All @@ -926,6 +951,11 @@ export const VertexAnthropicChatCompleteStreamChunkTransform: (
}

if (parsedChunk.type === 'message_delta' && parsedChunk.usage) {
const totalTokens =
(streamState?.usage?.prompt_tokens ?? 0) +
(streamState?.usage?.cache_creation_input_tokens ?? 0) +
(streamState?.usage?.cache_read_input_tokens ?? 0) +
(parsedChunk.usage.output_tokens ?? 0);
return (
`data: ${JSON.stringify({
id: fallbackId,
Expand All @@ -945,10 +975,8 @@ export const VertexAnthropicChatCompleteStreamChunkTransform: (
],
usage: {
completion_tokens: parsedChunk.usage?.output_tokens,
prompt_tokens: streamState.usage?.prompt_tokens,
total_tokens:
(streamState.usage?.prompt_tokens || 0) +
(parsedChunk.usage?.output_tokens || 0),
...streamState.usage,
total_tokens: totalTokens,
},
})}` + '\n\n'
);
Expand Down