DeepSeek released 3 new open-source V4 models
DeepSeek releases V4 series with 1M-token context and Sparse Attention as open weights, targeting open-source SOTA.
Excerpt
DeepSeek unveils the V4 series with a million-token context, new Sparse Attention, and open weights, aiming for open-source SOTA performance.
Read at source: https://www.testingcatalog.com/deepseek-released-3-new-open-source-v4-models/