agent/codex
2
0
Fork 0
mirror of https://github.com/pchuan98/codex.git synced 2026-07-01 00:31:56 +08:00
Code Issues Packages Projects Releases Wiki Activity

Add per-session realtime model and version overrides (#24999)

Browse Source 
## Why

Clients need to select a realtime session configuration for an
individual start without rewriting persisted configuration or restarting
the app-server process.

## What Changed

- Add optional `model` and `version` fields to `thread/realtime/start`
- Forward those optional values through the realtime start operation and
apply them only for that session
- Preserve existing configured/default behavior when the new fields are
omitted
- Update generated protocol schema and app-server documentation

## Validation

- Added/updated protocol serialization coverage for the new optional
request fields
- Added focused core coverage for a session override taking precedence
over configured realtime selection
- Added focused app-server coverage that a request override reaches the
realtime WebSocket handshake
This commit is contained in:
guinness-oai
2026-06-09 17:54:32 -07:00
committed by GitHub
parent 5ac640ac49
commit 4a3eac2144
11 changed files with 145 additions and 13 deletions
Download Patch File Download Diff File Expand all files Collapse all files
codex-rs/app-server/tests/suite/v2/experimental_api.rs
+4
View File
@@ -80,10 +80,12 @@ async fn realtime_conversation_start_requires_experimental_api_capability() -> R
let request_id = mcp
.send_thread_realtime_start_request(ThreadRealtimeStartParams {
thread_id: "thr_123".to_string(),
model: None,
output_modality: RealtimeOutputModality::Audio,
prompt: Some(Some("hello".to_string())),
realtime_session_id: None,
transport: None,
version: None,
voice: None,
})
.await?;
@@ -186,12 +188,14 @@ async fn realtime_webrtc_start_requires_experimental_api_capability() -> Result<
let request_id = mcp
.send_thread_realtime_start_request(ThreadRealtimeStartParams {
thread_id: "thr_123".to_string(),
model: None,
output_modality: RealtimeOutputModality::Audio,
prompt: Some(Some("hello".to_string())),
realtime_session_id: None,
transport: Some(ThreadRealtimeStartTransport::Webrtc {
sdp: "v=offer\r\n".to_string(),
}),
version: None,
voice: None,
})
.await?;
codex-rs/app-server/tests/suite/v2/realtime_conversation.rs
+18
View File
@@ -311,12 +311,14 @@ impl RealtimeE2eHarness {
.mcp
.send_thread_realtime_start_request(ThreadRealtimeStartParams {
thread_id: self.thread_id.clone(),
model: None,
output_modality: RealtimeOutputModality::Audio,
prompt: Some(Some("backend prompt".to_string())),
realtime_session_id: None,
transport: Some(ThreadRealtimeStartTransport::Webrtc {
sdp: offer_sdp.to_string(),
}),
version: None,
voice: None,
})
.await?;
@@ -556,10 +558,12 @@ async fn realtime_conversation_streams_v2_notifications() -> Result<()> {
let start_request_id = mcp
.send_thread_realtime_start_request(ThreadRealtimeStartParams {
thread_id: thread_start.thread.id.clone(),
model: Some("realtime-treatment-model".to_string()),
output_modality: RealtimeOutputModality::Audio,
prompt: None,
realtime_session_id: None,
transport: None,
version: None,
voice: Some(RealtimeVoice::Cedar),
})
.await?;
@@ -588,6 +592,10 @@ async fn realtime_conversation_streams_v2_notifications() -> Result<()> {
startup_context_request.body_json()["session"]["audio"]["output"]["voice"],
"cedar"
);
assert_eq!(
realtime_server.single_handshake().uri(),
"/v1/realtime?model=realtime-treatment-model"
);
assert_eq!(
startup_context_request.body_json()["session"]["output_modalities"],
json!(["audio"])
@@ -805,10 +813,12 @@ async fn realtime_text_output_modality_requests_text_output_and_final_transcript
let start_request_id = mcp
.send_thread_realtime_start_request(ThreadRealtimeStartParams {
thread_id: thread_start.thread.id.clone(),
model: None,
output_modality: RealtimeOutputModality::Text,
prompt: None,
realtime_session_id: None,
transport: None,
version: None,
voice: None,
})
.await?;
@@ -979,10 +989,12 @@ async fn realtime_conversation_stop_emits_closed_notification() -> Result<()> {
let start_request_id = mcp
.send_thread_realtime_start_request(ThreadRealtimeStartParams {
thread_id: thread_start.thread.id.clone(),
model: None,
output_modality: RealtimeOutputModality::Audio,
prompt: Some(Some("backend prompt".to_string())),
realtime_session_id: None,
transport: None,
version: None,
voice: None,
})
.await?;
@@ -1076,12 +1088,14 @@ async fn realtime_webrtc_start_emits_sdp_notification() -> Result<()> {
let start_request_id = mcp
.send_thread_realtime_start_request(ThreadRealtimeStartParams {
thread_id: thread_id.clone(),
model: None,
output_modality: RealtimeOutputModality::Audio,
prompt: Some(Some("backend prompt".to_string())),
realtime_session_id: None,
transport: Some(ThreadRealtimeStartTransport::Webrtc {
sdp: "v=offer\r\n".to_string(),
}),
version: None,
voice: None,
})
.await?;
@@ -1991,12 +2005,14 @@ async fn realtime_webrtc_start_surfaces_backend_error() -> Result<()> {
let start_request_id = mcp
.send_thread_realtime_start_request(ThreadRealtimeStartParams {
thread_id: thread_start.thread.id,
model: None,
output_modality: RealtimeOutputModality::Audio,
prompt: Some(Some("backend prompt".to_string())),
realtime_session_id: None,
transport: Some(ThreadRealtimeStartTransport::Webrtc {
sdp: "v=offer\r\n".to_string(),
}),
version: None,
voice: None,
})
.await?;
@@ -2050,10 +2066,12 @@ async fn realtime_conversation_requires_feature_flag() -> Result<()> {
let start_request_id = mcp
.send_thread_realtime_start_request(ThreadRealtimeStartParams {
thread_id: thread_start.thread.id.clone(),
model: None,
output_modality: RealtimeOutputModality::Audio,
prompt: Some(Some("backend prompt".to_string())),
realtime_session_id: None,
transport: None,
version: None,
voice: None,
})
.await?;