Skip to content

fix(auto-routing-benchmark): bill BYOK sweep completions at upstream cost #9000

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

Merged
iscekic merged 1 commit into from
Oct 28, 2026
Merged

fix(auto-routing-benchmark): bill BYOK sweep completions at upstream cost #9000

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
19 changes: 19 additions & 0 deletions services/auto-routing-benchmark/src/scoreboard/sweep.test.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -247,6 +247,25 @@ describe('readSweepCompletion', () => {
});
});

it.each([
[
'the provider-billed upstream cost for BYOK',
{ cost: 0.0004, is_byok: true, cost_details: { upstream_inference_cost: 0.008 } },
0.008,
],
[
'the router cost without BYOK',
{ cost: 0.012, is_byok: false, cost_details: { upstream_inference_cost: 0.009 } },
0.012,
],
['an unknown cost when BYOK omits the upstream cost', { cost: 0.0004, is_byok: true }, null],
])('reports %s', async (_label, usage, expected) => {
const response = new Response(
`data: {"choices":[{"delta":{"content":"answer"}}]}\n\ndata: ${JSON.stringify({ choices: [], usage })}\n\ndata: [DONE]\n\n`
);
expect((await readSweepCompletion(response, 100, () => 175)).costUsd).toBe(expected);
});

it('accepts explicit null error fields in answer and usage chunks', async () => {
const response = new Response(
[
Expand Down
13 changes: 12 additions & 1 deletion services/auto-routing-benchmark/src/scoreboard/sweep.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -32,6 +32,10 @@ const streamChunkSchema = z.object({
usage: z
.object({
cost: z.number().nonnegative().nullish(),
is_byok: z.boolean().nullish(),
cost_details: z
.object({ upstream_inference_cost: z.number().nonnegative().nullish() })
.nullish(),
prompt_tokens: z.number().int().nonnegative().nullish(),
completion_tokens: z.number().int().nonnegative().nullish(),
})
Expand Down Expand Up @@ -83,7 +87,14 @@ export async function readSweepCompletion(
}
output += delta.content ?? '';
}
if (chunk.usage?.cost != null) costUsd = chunk.usage.cost;
if (chunk.usage) {
// With BYOK, `cost` is only OpenRouter's fee; the provider bills the upstream inference cost.
const billed =
chunk.usage.is_byok === true
? chunk.usage.cost_details?.upstream_inference_cost
: chunk.usage.cost;
if (billed != null) costUsd = billed;
}
if (chunk.usage?.prompt_tokens != null || chunk.usage?.completion_tokens != null) {
tokenUsage = {
promptTokens: chunk.usage.prompt_tokens ?? tokenUsage?.promptTokens ?? null,
Expand Down