Ling 3.1 Flash Is Free to Try; Its 560B Weights Are Not Public
Ling 3.1 Flash is available through hosted APIs, with 560 billion total parameters and about 25 billion activated per token….
Klaus Fischer Editorial profile
EyesTech editorial profile covering open models and control. Articles distinguish source reporting, analysis and documented tests. EyesTech uses AI assistance; this coverage profile is not a verified professional résumé. Corrections: https://eyestech.in/contact/
Holo4: What Open Computer-Use Agents Can Actually Automate
Holo4 combines GUI, code and tool use. Examine its 41.5% full-task success rate, model licenses and the checks needed before automating real workflows.
DeepSeek Ascend: Fast Kernels, Limited Reproducibility
DeepSeek has published Ascend 950 versions of two pieces of its model infrastructure: DeepEP-Ascend for moving mixture-of-experts tokens between devices,…
Huawei openPangu-2.0 Opens Its Ascend Training Stack
Huawei opens pretraining, SFT and RL code for openPangu-2.0. Explore the 92B examples, Ascend dependencies and what missing datasets mean for developers.
The NixOS Sovereign Cloud: Why Dutch Gov Ditched Microsoft
The Dutch Ministry of the Interior (MinBZK) has officially initiated a nationwide public-sector migration away from Microsoft Windows and Microsoft…
DeepSeek MLA Architecture: How It Cuts KV Cache by 93%
DeepSeek Multi-Head Latent Attention compresses KV cache down to 0.16 KB/token/layer. Here is the low-rank projection math and 128k context serving economics.