DeepSeek Deepseek-v4.1-flash-v2 coming 👀!! I'm more convinced something has changed behind the deepseek-platform api. The "deepseek-flash" endpoint seems to be routing to a different model now. It could be: - Deepseek-v4.2-flash - An updated checkpoint of v4.1-flash Since it's also a little slower, i wonder if V4.1-pro is being... quietly tested behind the endpoint. But either way, i'm getting way better result, the difference is noticeable. See the comparison video, same prompt, same settings.show more

@studio_yebisu DeepSeek側で絶対何か変わった気がする！

@deepseek_ai Sorry but the same brother. Tested some things too but the model is just random as fuck

@deepseek_ai I'm not ruling out randomness, but to me difference is there, so i'm updating my bench to take this into account. Need to get more data.

@deepseek_ai This is so interesting, can’t wait to try it

@deepseek_ai We are expecting another one soon, it's about time.

@deepseek_ai keeping one fixed prompt around catches swaps like this

@deepseek_ai Ya, prompts and settings are fixed, i have versioning as well.

@deepseek_ai Looks same

@deepseek_ai 昨天我也感觉api的性能强了很多，我让他执行了一系列的任务，交付结果非常好

@deepseek_ai so where is my v4.1pro

@deepseek_ai no bro ,this is impossible, because Genshin Impact's next update is on November 4th

@deepseek_ai If it's DeepSeek-V4.1 Pro, I'm pretty disappointed too

@deepseek_ai Hmm... interesting. Is this just the Deepseek Flash endpoint or the Deepseek v4.1 Flash Endpoint as well?

@deepseek_ai I have the same experience, my Deepseek has been so thorough and cautious over the last week, almost feel like Opus sometimes.

@deepseek_ai HEHEHEHEHEHEHEH <3 Oh man if we get DS V4.1 Flash 1010 I will be so happy!! <3

@deepseek_ai This

@deepseek_ai 毫无感觉到变化

@deepseek_ai 我个人猜测可能是deepseek mini，一款基于4.1f架构的，能跑在5090等级的消费卡上，且输出速度接近100tok/s的模型。因为世界知识貌似比4.1f更差，而某些方面更强

@deepseek_ai I also noticed it. It got very good at computer use and UI testing.

@deepseek_ai 哪裡可以用？

@deepseek_ai I'm calling direct api.