electricjimi

Results 5 comments of electricjimi

No I don't stream responses, however I am using the Async client and not the sync one. Could that be the cause of the issue?

Sure! ```python import asyncio from pprint import pprint from dotenv import load_dotenv from langfuse.openai import AsyncOpenAI load_dotenv() oai_client = AsyncOpenAI() async def analyze_image(base64_img: str): system_prompt = "You are an image...

I see token counts are correct, but the problem may lie in how the pricing is calculated by openai, img tokens have a different price.

Any news on this? Or any workarounds for the moment? Thank you.