TheLovinator/ANewDawn
1
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Refactor OpenAI response handling to use updated response types

Browse source
This commit is contained in:
Joakim Hellsén 2025-08-17 04:38:33 +02:00
commit 7a6b4d8fce
1 changed files with 4 additions and 4 deletions
Download patch file Download diff file Expand all files Collapse all files

8
misc.py
View file

@ -15,7 +15,7 @@ if TYPE_CHECKING:
from discord.guild import GuildChannel from discord.guild import GuildChannel
from discord.interactions import InteractionChannel from discord.interactions import InteractionChannel
from openai import OpenAI from openai import OpenAI
from openai.types.chat.chat_completion import ChatCompletion from openai.types.responses import Response
logger: logging.Logger = logging.getLogger(__name__) logger: logging.Logger = logging.getLogger(__name__)
@ -194,11 +194,11 @@ def chat( # noqa: PLR0913, PLR0917
"User message ends here.\n" "User message ends here.\n"
) )
completion: ChatCompletion = openai_client.chat.completions.create( resp: Response = openai_client.responses.create(
model="gpt-5-chat-latest", model="gpt-5-chat-latest",
messages=[{"role": "system", "content": prompt}], input=[{"role": "user", "content": prompt}],
) )
response: str | None = completion.choices[0].message.content response: str | None = resp.output_text
logger.info("AI response: %s", response) logger.info("AI response: %s", response)
return response return response