Not sure if it’s because they caught abusers, but now slow requests are just “slow” instead of “slow slow slow slow slow” requests. It’s bearable now… Is everyone experiencing the same? I’m curious!
2 Likes
I feel it too! I might be biased because my solution for slow mode is multitasking (switching windows constantly and working on multiple projects instead of waiting for the response), but at least for gemini-pro-2.5 is much better. For claude… it seems 10% faster, when it doesn’t block me because the pool is too large.
lol, it is still slow for me.
How are you guys burning through the 500+ requests a month? I’m trying hard but didn’t even get beyond 380 requests when avoiding smallish/non-value requests
1 Like
You need to work on complex things to burn it. like atomicity, concurrency, race conditions..
nothing different. very slow for every prompt. tried 3.7 and gemini 2.5, stuck on generating.
May 1, 2025, 03:26 PM | Agent | claude-3.7-sonnet-thinking-max | Errored, Not Charged | 1 |
---|---|---|---|---|
May 1, 2025, 03:24 PM | Agent | gpt-4o | Errored, Not Charged | 1 |
May 1, 2025, 03:23 PM | Agent | gemini-2.5-pro-exp-03-25 | Errored, Not Charged | 1 |
May 1, 2025, 03:21 PM | Agent | gpt-4.1 | Errored, Not Charged | 1 |
May 1, 2025, 03:18 PM | Agent | gemini-2.5-pro-exp-03-25 | Errored, Not Charged | 1 |
May 1, 2025, 03:11 PM | Agent | claude-3.7-sonnet-thinking | Errored, Not Charged | 2 |