Atharva Ingle(@AtharvaIngle7):I've been using DeepSeek V4.1 Flash 1017, and it's so good, much better than GLM 5.3 / GLM 5.3 Flash. Feels more reliable and is so fast that it really helps me stay focused during in-the-loop work. It's already live on our Token Factory at @jarvislabsai, and we're actively working on making it even faster (currently at 180–200 tok/s). PS: The following video is not sped up 👀

Atharva Ingle

@AtharvaIngle7

AI @jarvislabsai , @kaggle Competition Expert

Joined November 2020

329 Following 2.3K Followers

Atharva Ingle@AtharvaIngle7

2026.10.24 11:34

I've been using DeepSeek V4.1 Flash 1017, and it's so good, much better than GLM 5.3 / GLM 5.3 Flash. Feels more reliable and is so fast that it really helps me stay focused during in-the-loop work. It's already live on our Token Factory at @jarvislabsai, and we're actively working on making it even faster (currently at 180–200 tok/s). PS: The following video is not sped up 👀