// event · sdk
openai-python 3.15.0: agent session settings, managed Responses WebSockets, prompt-cache prewarming
sdkvendor: openaipublished: 25d agoconfidence: 0.90id: 90c90225-628e-47bb-8a73-7a9842324ef0
OpenAI Python SDK v3.15.0 adds agent session model settings, audio-mini model choices, compaction progress events, managed Responses WebSocket sessions, and prompt-cache prewarming. Also preserves chat stream moderation results and clarifies SIP call ID usage for realtime voice agents.
capability delta
an agent can now configure per-session model settings, receive compaction progress events, use managed Responses WebSocket sessions, and prewarm prompt caches via the OpenAI Python SDK.
// install
pip install openai==3.15.0pypi · openai @ 3.15.0
// example
python
from openai import OpenAI
client = OpenAI()
# Managed Responses WebSocket session with agent session model settings
with client.responses.sessions.connect(
model="gpt-5",
session_settings={
"prompt_cache_prewarm": True,
"compaction": {"emit_progress": True},
},
) as session:
session.send({"role": "user", "content": "Plan my day."})
for event in session:
if event.type == "compaction.progress":
print("compaction:", event.progress)
elif event.type == "response.output_text.delta":
print(event.delta, end="")
// links