What are the top AI Models that are still relevant today from 2024 and 2025?
In 2026, the AI landscape has seen a staggering evolution, with Mixture of Experts (MoE) models becoming prevalent. Dense models are largely obsolete, though some exceptions remain. The discussion highlights the ability to handle large parameter models, indicating a preference for powerful AI systems. The Kimi k3 model is specifically mentioned as being out of contention for large quantization.
This discussion, unlike typical forward-looking analyses, reflects on the relevance of 2024 and 2025 AI models from the perspective of 2026, highlighting the shift to MoE models.
时间与来源
时间显示为 UTC
显示时区：UTC
本地时区尚不可用，暂时显示 UTC。
发布当时偏移：UTC+02026年9月13日 12:23 UTC
收录当时偏移：UTC+02026年9月13日 15:01 UTC
- 发布
- 2026年9月13日 12:23
- 收录
- 2026年9月13日 15:01
- 来源类型
- 开发者社区
- 档位
- 社区
- 信源状态
- 正常
档位是按信源手工设定的编辑判断，不是逐条打分。
We are in 2026 and it has been a crazy year. The evolution of the technology has been staggering to say the least. Pretty much we are on the MoE period and dense models are almost in the way of the dodo except for a few.
I was just thinking, are there any models of the last 2 years that you would fire up and still feel useful? Example Deep Seek R1.
If you have recommendations write them down. I have 28TB of storage and I am backing up relevant models that can still be useful, from current to older ones. But clearly not all are worth keeping.
Noted: I can hold large parameter models therefore not limited to small ones. No Kimi k3 at large quant thats out of the question lol