github/codex
4
0
Fork 0
mirror of https://github.com/openai/codex.git synced 2026-10-19 16:17:49 +00:00
Code Issues Packages Projects Releases Wiki Activity

Preserve native USD amounts in thread usage responses (#50442)

Browse Source 
## What changed

Add optional `native_usage_usd_micros` fields to thread usage totals and breakdown groups, and carry backend values through the client alongside existing credit and USD estimates. Missing or null native amounts remain `None`, while explicit zero amounts remain `Some(0)`.

## Testing

Add coverage for native totals and group amounts, mixed batches with legacy responses, and the distinction between missing, null, and zero native group amounts.

GitOrigin-RevId: 53891fc95807138e2eccb99d4cda1520b857dba4
This commit is contained in:
asohrabi-oai
2026-10-02 22:20:55 +00:00
committed by copyberry
parent 0df76892b1
commit 3629508849
4 changed files with 172 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files
codex-rs/backend-client/src/client/thread_usage.rs
+6 -2
View File
@@ -1,4 +1,4 @@
//! Authoritative estimated credit and dollar usage for bounded batches of Codex threads.
//! Backend credit estimates and native-dollar usage for bounded batches of Codex threads.
use super::Client;
use super::PathStyle;
@@ -17,6 +17,7 @@ pub struct ThreadUsageBreakdownGroup {
pub reasoning_effort: Option<String>,
pub speed: Option<String>,
pub estimated_usage_credits_micros: i64,
pub native_usage_usd_micros: Option<i64>,
pub net_new_input_tokens: Option<i64>,
pub cached_input_tokens: Option<i64>,
pub input_tokens: Option<i64>,
@@ -24,12 +25,13 @@ pub struct ThreadUsageBreakdownGroup {
pub total_tokens: Option<i64>,
}
/// Backend-estimated usage totals expressed in integer millionths.
/// Backend usage totals expressed in integer millionths of their respective units.
#[derive(Clone, Debug, Deserialize, PartialEq, Eq)]
pub struct ThreadUsage {
pub thread_id: String,
pub estimated_usage_credits_micros: i64,
pub estimated_usage_usd_micros: Option<i64>,
pub native_usage_usd_micros: Option<i64>,
#[serde(default)]
pub groups: Vec<ThreadUsageBreakdownGroup>,
}
@@ -50,6 +52,7 @@ struct ThreadUsageEstimate {
thread_id: String,
estimated_usage_credits_micros: Option<i64>,
estimated_usage_usd_micros: Option<i64>,
native_usage_usd_micros: Option<i64>,
groups: Option<Vec<ThreadUsageBreakdownGroup>>,
}
@@ -103,6 +106,7 @@ impl Client {
thread_id: row.thread_id,
estimated_usage_credits_micros: row.estimated_usage_credits_micros?,
estimated_usage_usd_micros: row.estimated_usage_usd_micros,
native_usage_usd_micros: row.native_usage_usd_micros,
groups: row.groups.unwrap_or_default(),
})
})
codex-rs/backend-client/src/client/thread_usage_tests.rs
+163
View File
@@ -41,11 +41,13 @@ async fn get_thread_usage_returns_requested_thread_totals() {
"thread_id": "thread-123",
"estimated_usage_credits_micros": 46_000_000,
"estimated_usage_usd_micros": 1_820_000,
"native_usage_usd_micros": 3_250_000,
"groups": [{
"model": "gpt-5.4",
"reasoning_effort": "high",
"speed": "fast",
"estimated_usage_credits_micros": 46_000_000,
"native_usage_usd_micros": 3_250_000,
"net_new_input_tokens": 80,
"cached_input_tokens": 20,
"input_tokens": 100,
@@ -71,11 +73,13 @@ async fn get_thread_usage_returns_requested_thread_totals() {
thread_id: "thread-123".to_string(),
estimated_usage_credits_micros: 46_000_000,
estimated_usage_usd_micros: Some(1_820_000),
native_usage_usd_micros: Some(3_250_000),
groups: vec![ThreadUsageBreakdownGroup {
model: Some("gpt-5.4".to_string()),
reasoning_effort: Some("high".to_string()),
speed: Some("fast".to_string()),
estimated_usage_credits_micros: 46_000_000,
native_usage_usd_micros: Some(3_250_000),
net_new_input_tokens: Some(80),
cached_input_tokens: Some(20),
input_tokens: Some(100),
@@ -115,6 +119,7 @@ async fn get_thread_usage_accepts_credits_without_usd_estimate() {
thread_id: "thread-123".to_string(),
estimated_usage_credits_micros: 46_000_000,
estimated_usage_usd_micros: None,
native_usage_usd_micros: None,
groups: Vec::new(),
}
);
@@ -198,3 +203,161 @@ async fn batch_usage_rejects_duplicate_and_unrequested_response_rows() {
}
}
}
#[test]
fn thread_usage_preserves_unknown_and_zero_native_group_amounts() {
let usage: ThreadUsage = serde_json::from_value(json!({
"thread_id": "thread-123",
"estimated_usage_credits_micros": 0,
"native_usage_usd_micros": null,
"groups": [
{ "model": "legacy", "estimated_usage_credits_micros": 0 },
{
"model": "unknown",
"estimated_usage_credits_micros": 0,
"native_usage_usd_micros": null
},
{
"model": "zero",
"estimated_usage_credits_micros": 0,
"native_usage_usd_micros": 0
}
]
}))
.expect("decode native groups from legacy and current backend responses");
assert_eq!(
usage,
ThreadUsage {
thread_id: "thread-123".to_string(),
estimated_usage_credits_micros: 0,
estimated_usage_usd_micros: None,
native_usage_usd_micros: None,
groups: [("legacy", None), ("unknown", None), ("zero", Some(0))]
.into_iter()
.map(
|(model, native_usage_usd_micros)| ThreadUsageBreakdownGroup {
model: Some(model.to_string()),
reasoning_effort: None,
speed: None,
estimated_usage_credits_micros: 0,
native_usage_usd_micros,
net_new_input_tokens: None,
cached_input_tokens: None,
input_tokens: None,
output_tokens: None,
total_tokens: None,
}
)
.collect(),
}
);
}
#[tokio::test]
async fn get_threads_usage_preserves_native_amounts_and_absence() {
let server = MockServer::start().await;
Mock::given(method("POST"))
.and(path("/api/codex/usage/thread_usage/query"))
.and(body_json(json!({ "thread_ids": ["priced", "unknown"] })))
.respond_with(ResponseTemplate::new(/*s*/ 200).set_body_json(json!({
"threads": [
{
"thread_id": "priced",
"estimated_usage_credits_micros": 46_000_000,
"estimated_usage_usd_micros": 2_505_000,
"native_usage_usd_micros": 3_250_000,
"groups": [
{
"model": "test-model-a",
"reasoning_effort": "medium",
"speed": "standard",
"estimated_usage_credits_micros": 12_000_000,
"native_usage_usd_micros": 750_000
},
{
"model": "test-model-b",
"reasoning_effort": "high",
"speed": "fast",
"estimated_usage_credits_micros": 34_000_000,
"native_usage_usd_micros": 2_500_000
}
]
},
{
"thread_id": "unknown",
"estimated_usage_credits_micros": 7_000_000,
"groups": [{
"model": "legacy-model",
"estimated_usage_credits_micros": 7_000_000
}]
}
]
})))
.expect(/*r*/ 1)
.mount(&server)
.await;
let client = Client::new(
server.uri(),
HttpClientFactory::new(OutboundProxyPolicy::ReqwestDefault),
);
assert_eq!(
client
.get_threads_usage(&["priced", "unknown"])
.await
.expect("read native and legacy usage in one batch"),
vec![
ThreadUsage {
thread_id: "priced".to_string(),
estimated_usage_credits_micros: 46_000_000,
estimated_usage_usd_micros: Some(2_505_000),
native_usage_usd_micros: Some(3_250_000),
groups: vec![
ThreadUsageBreakdownGroup {
model: Some("test-model-a".to_string()),
reasoning_effort: Some("medium".to_string()),
speed: Some("standard".to_string()),
estimated_usage_credits_micros: 12_000_000,
native_usage_usd_micros: Some(750_000),
net_new_input_tokens: None,
cached_input_tokens: None,
input_tokens: None,
output_tokens: None,
total_tokens: None,
},
ThreadUsageBreakdownGroup {
model: Some("test-model-b".to_string()),
reasoning_effort: Some("high".to_string()),
speed: Some("fast".to_string()),
estimated_usage_credits_micros: 34_000_000,
native_usage_usd_micros: Some(2_500_000),
net_new_input_tokens: None,
cached_input_tokens: None,
input_tokens: None,
output_tokens: None,
total_tokens: None,
},
],
},
ThreadUsage {
thread_id: "unknown".to_string(),
estimated_usage_credits_micros: 7_000_000,
estimated_usage_usd_micros: None,
native_usage_usd_micros: None,
groups: vec![ThreadUsageBreakdownGroup {
model: Some("legacy-model".to_string()),
reasoning_effort: None,
speed: None,
estimated_usage_credits_micros: 7_000_000,
native_usage_usd_micros: None,
net_new_input_tokens: None,
cached_input_tokens: None,
input_tokens: None,
output_tokens: None,
total_tokens: None,
}],
},
]
);
}