MDL
DeepSeek-V2-Lite
deepseek-ai
Capability
—
percentile index
Downloads
448.8K
on HF
Likes
176
Avg score
—
Weights
Open
Released
May 15, 2024
About
DeepSeek-V2-Lite is a open-weights model in the DeepSeek family from deepseek-ai.
othertext-generation
Benchmark scores · 0
No benchmark scores yet.
The research behind this model
all papers mentioning it →DeepSeek-V2: A Strong, Economical, and Efficient Mixture-of-Experts Language ModelPruning and Distilling Mixture-of-Experts into Dense Language Models▲ 1 on HF · May 27, 2026Diagnosing Overhead in Dispatch Operations: Cross-architecture ObservatoryMay 20, 2026When Are Experts Misrouted? Counterfactual Routing Analysis in Mixture-of-Experts Language ModelsMay 8, 2026Irminsul: MLA-Native Position-Independent Caching for Agentic LLM ServingMay 7, 2026
Mentions matched by name in title/abstract — from the arXiv + HF daily corpus.