API Timeout on Tier 5 using gpt-5 with reasoning effort low

Hello, it’s been 2 weeks since I started getting a large number of API timeouts when using chat.completions.createfrom AsyncOpenAI inside one of our microservices. I’m using GPT-5 with the reasoning effort set to low.

I tried changing the API key, as some users mentioned this as a potential solution in other forum posts, but it didn’t make any difference.

What could be causing this type of error?