Hey everyone, I’ve been using DeepSeek Chat for a few months now to help me draft and debug my data pipeline SQL, especially for BigQuery and dbt. I’m still pretty new to all this, so I rely on it a lot for quick explanations and pattern suggestions.
After the last update (I think it was around mid-week?), I’ve noticed the responses are taking noticeably longer to come back. I’m talking about a delay of several seconds before it even starts generating, where before it felt almost instant. It’s not my internet—my other tools are fine.
I ran a super basic test this morning. I asked the same question five times: “Write a SQL CTE to deduplicate a table based on the latest timestamp.” I timed from hitting enter to seeing the first token. My average went from about 1.2 seconds last week to around 3.8 seconds now. That’s like a 3x increase 😅.
Has anyone else experienced this? I’m wondering if it’s something on my end, a regional thing, or if the update changed how the models are served. It’s still super helpful when it replies, but the wait is starting to interrupt my flow when I’m iterating on a complex transformation logic.
Oh wow, thanks for actually timing it. I hadn't thought to do that, but I definitely feel the same wait. I use it for almost the same thing, piecing together incremental models in dbt, and that initial lag really throws off my rhythm when I'm trying to think.
I was worried it was just me, maybe my local setup or something with the CLI. Good to know it's not in my head. Have you tried at different times of day? I sometimes wonder if it's a load thing, like more people using it after an update.