I have been seeing this as well, started on Monday. Was fine the week prior. Many requests to the bars endpoint failing and needing a retry, those that come through are at a very high latency.
Has been bad enough for me to stop trading for a bit while I work out a websockets input, which is a more efficient solution in my situation anyway.
Yes thats about right it started on Monday, I’ve increased my timeouts to 4-5 seconds but still seeing timeouts. Hopefully someone will fix the problem.
I have the same issue. Connection to data.alpaca.markets/v2/stocks/bars fails for multiple ticker symbols. I am using python requests. Requests library throws a requests.exceptions.ConnectionError. Timeout and max retires exceeded are common error messages. Happens to me about once every 20 requests. Issue started on April 13, 2026. My code has not changed in months. Tried increasing timeouts to 30 seconds, no change.
I’m seeing this as well. Since the scheduled maintenance on Saturday, April 11th, I’ve seen a lot of ECONNRESET responses with the Data API (across multiple endpoints). And my response times have gone from an average of less than 100ms to averaging between 300ms to 400ms depending on the day.
And I agree with the original post, the symptoms seem very similar to what was happening a month or two ago when there was a similar issue.
I’ve cancelled my subscription. 8 days after this issue was brought up, it’s not even clear that anyone at Alpaca is working on this. Good luck to all,
From what I can tell it looks like a load issue, outside market hours before 9:30 and after 4:00pm EST the access times drop to normal ranges. It looks like Alpaca api endpoint is using a google cloud compute VM, so maybe it is getting overloaded or is under sized?
This is happening for me too. This is not the first time this is an issue either, last time was around the start of February and the call did not work around market open (on and off) for about a week or two.
@dcbz Could you provide an example of a request that seems to take a long time? About what time was the request? I could check the logs and see if anything stands out as being odd.
both API calls of type below, now after 4pm EST there is no delay, now they are fast (20-40ms), no timeouts or delays as happening during the market hours 9:30-4pm EST . This looks like the same thing that was happening about 4 weeks ago which cleared up about 2 weeks ago, it may be back now, same latencies. I will recheck tommorow during market hours and recheck my machine.
The issue is back for me too. Seemed to have been solved two weeks ago like @dcbz said. My issues began again on the morning of May 26 and have persisted since. Same endpoints as @dcbz , requests timing out.