DeepSeek V4 Flash (IQ2_XXS)

antirez's 2-bit community GGUF of DeepSeek V4 Flash 0731 — the July 2026 re-post-trained checkpoint of the 284B-parameter Mixture-of-Experts (~13B active per token) with 1M-token context and Compressed Sparse Attention. Same architecture as the original V4 Flash, with substantially stronger agentic and coding behaviour. The routed experts are crushed to IQ2_XXS / Q2_K while attention projections, shared experts, router, embeddings, and output head stay at Q8_0 / F16, landing the whole model in ~81 GB — runnable on 128 GB-class machines versus ~156 GB for the native FP4 build. Note: these quants were produced for antirez's ds4 inference engine; whether they load on stock llama depends on its DeepSeek-V4 support.

Gezel downloads this model the first time a gezel needs it, checks the file against its published hash, and applies tuned settings for it.

The model’s own page · Every model · Model scorecard