HF Model & Benchmark Matrix

DeepSeek-V4-Flash-Base

deepseek-ai/DeepSeek-V4-Flash-Base

kind

fp8

quant

fp8 (8-bit)

family

deepseek

modality

—

root model

deepseek-ai/DeepSeek-V4-Flash

provider

deepseek-ai

license

—

language

—

context

—

library

—

vRAM (GB)

—

size on disk

—

downloads

205,182

likes

283

created

—

last modified

2026-04-27T06:51:43+00:00 (2026)

huggingface

https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-Base

tags

safetensorsdeepseek_v4fp8region:us

Benchmark scores (0)

no scores extracted for this variant.

Discussions (0)

none scraped.