DeepSeek-V4-Flash-Vision-Exp

Model tree for deepseek-ai/DeepSeek-V4-Flash-Vision-Exp

Finetunes

10 models

Merges

1 model

Quantizations

25 models

Spaces using deepseek-ai/DeepSeek-V4-Flash-Vision-Exp 3

Collection including deepseek-ai/DeepSeek-V4-Flash-Vision-Exp

[

DeepSeek-V4

Collection

10 items • Updated 1 day ago • 900

](https://huggingface.co/collections/deepseek-ai/deepseek-v4)

License

This repository is licensed under the MIT License.

Safetensors

Model size

305B params

Tensor type

BF16

·

F32

·

F8_E4M3

·

I8

·

I64

·

How to Run with vLLM

For example, the command below serves the model with vLLM on a single 4×GB300 node. See the vLLM recipe for detailed instructions and other hardware configurations.

docker run --gpus all \
  vllm/vllm-openai:deepseekv4-flash-vision deepseek-ai/DeepSeek-V4-Flash-Vision-Exp \
  --kv-cache-dtype fp8 \
  --block-size 256 \
  --tensor-parallel-size 4 \
  --tool-call-parser deepseek_v4 \
  --enable-auto-tool-choice \
  --reasoning-parser deepseek_v4 \
  --reasoning-config '{"reasoning_parser":"deepseek_v4","reasoning_start_str":"","reasoning_end_str":""}' \
  --speculative-config '{"method":"dspark","model":"deepseek-ai/DeepSeek-V4-Flash-Vision-Exp","num_speculative_tokens":3,"draft_sample_method":"probabilistic","enable_adaptive_verification":true}'

How to Run with SGLang

Enable DSpark with --speculative-algorithm DSPARK and do not set a separate --speculative-draft-model-path as the target and draft weights therefore come from the same checkpoint. See the SGLang cookbook for detailed instructions, benchmarks and other hardwares configurations.

sglang serve \
  --model-path deepseek-ai/DeepSeek-V4-Flash-Vision-Exp \
  --tp 4 \
  --speculative-algorithm DSPARK \
  --mem-fraction-static 0.85 \
  --host 0.0.0.0 \
  --port 30000

Minimal inference

See inference/README.md for dependency installation, checkpoint conversion, and TXT/JSON inference commands.

Prompt encoding

See encoding/README.md. Both OpenAI-style JSON content blocks and the compact <image>path</image> TXT notation are supported. The two examples under inference/examples/ encode to identical prompts and token IDs.

Repository layout

This repository contains the tokenizer, prompt encoding reference, and a minimal PyTorch inference implementation for DeepSeek-V4 Flash Vision. The reference inference covers the vision encoder and aligner, DFlash attention, MoE, Hyper-Connections, and the DSpark forward path.

.
├── encoding/                  # OpenAI-style messages -> model prompt
├── inference/                 # weight conversion and minimal inference
│   └── examples/              # equivalent TXT and JSON vision prompts
├── config.json                # Hugging Face model metadata
├── generation_config.json
├── model.safetensors.index.json
├── tokenizer.json
└── tokenizer_config.json

encoding/ and inference/ deliberately remain separate: prompt formatting does not depend on PyTorch, while inference imports the sibling encoding module with an explicit Python path. No symlinks are required.

The tokenizer files are regular files so that the repository can be uploaded to Hugging Face without relying on local filesystem symlinks. The large model shards are described by model.safetensors.index.json and are not duplicated inside the source checkout used to assemble this repository.

Introduction

We are excited to introduce DeepSeek-V4-Flash-Vision-Exp, our first experimental multimodal model in the DeepSeek-V4 family. It builds on the DeepSeek-V4-Flash architecture by incorporating visual modules and undergoing continued training to unlock visual understanding capabilities.

Compared to DeepSeek-V4-Flash-0731, DeepSeek-V4-Flash-Vision-Exp achieves substantial improvements on its multimodal agent capabilities, while maintaining comparable performance on text-only agent tasks.

Notes:

1. For the text agent benchmarks above, DeepSeek models are evaluated with the minimal mode of DeepSeek Harness as the agent framework, using the max reasoning effort level with temperature = 1.0, top_p = 0.95.
2. † For ApexBench and Agents' Last Exam, DeepSeek-V4-Flash-0731 ignores the multimodal elements in the input.