DeepSeek V4 Pro

DeepSeek V4 Pro is a large-scale Mixture-of-Experts model from DeepSeek with 1.6T total parameters and 49B activated parameters, supporting a 1M-token context window. It is designed for advanced reasoning, coding, and long-horizon agent workflows, with strong performance across knowledge, math, and software engineering benchmarks. Built on the same architecture as DeepSeek V4 Flash, it introduces a hybrid attention system for efficient long-context processing. Reasoning efforts `high` and `xhigh` are supported; `xhigh` maps to max reasoning. It is well suited for complex workloads such as full-codebase analysis, multi-step automation, and large-scale information synthesis, where both capability and efficiency are critical.

Licence

mit

Architecture

DeepseekV4ForCausalLM

Parameters

1.60T

Context

1M tokens

Artifact format

safetensors

Quantization

fp8 8-bit

Download size

805 GB

Hardware to run it (inference, estimated)

1.7 TB VRAM minimum · 2.2 TB recommended · 2.2 TB system RAM · 806 GB storage

Estimated from parameter count and stored precision; verify against the selected runtime and context length.

Papers

DeepSeek-V4: Towards Highly Efficient Million-Token Context IntelligenceDeepSeek-AI, Anyi Xu, Bangcai Lin +316 more · 2026

Benchmarks

No benchmark observation recorded for this model.