Baidu Open-Sources ERNIE 5.5 Flash Base, a 280B-Parameter MoE Architecture
Baidu has open-sourced and released ERNIE 5.5 Flash Base, a MoE architecture model with 280B total parameters and 14B activated per token. Baidu Research claims it has reached SOTA at the same size and teases that the full version of ERNIE 5.5 Flash is coming soon. The reposter points out that its benchmark compares models of the same parameter scale rather than each company's latest versions, and lists DeepSeek-V4-Flash-Base corresponding to V4.1-Flash-1017, Hy3-preview-Base corresponding to Hy4, and MiMo-V2.5-Base corresponding to MiMo-V2.6-Flash for reference.
Why read this
AI score 67
Baidu has open-sourced a 280B/A14B MoE model; the author added comparisons with the latest versions of various companies at the same parameter scale, to help judge where this positioning actually stands.
Topics
View original postERNIE 5.5 Flash Base is now open source — a 280B-parameter model with 14B active per token, and it already matches SOTA at its size! And this is just the start — the full ERNIE 5.5 Flash is on its way. 👀
View original
ERNIE 5.5 Flash Base is now open source — a 280B-parameter model with 14B active per token, and it already matches SOTA at its size! And this is just the start — the full ERNIE 5.5 Flash is on its way. 👀
Body · AI translation
Baidu has open-sourced and released ERNIE 5.5 Flash Base
A MoE architecture model with 280B/A14B parameters
However, why are none of the models compared in the benchmark the latest models from each company? It turns out it's because they wanted to compare models of the same parameter scale. Alright, let's still list the latest models from each company for everyone's reference.
DeepSeek-V4-Flash-Base -> V4.1-Flash-1017 Hy3-preview-Base -> Hy4 MiMo-V2.5-Base -> MiMo-V2.6-Flash