Skip to model information

DeepSeek V4 Pro 0813

Weights available

DeepSeek V4 Pro 0813 weights published by DeepSeek in its Hugging Face organization (deepseek-ai/DeepSeek-V4-Pro-0813). Self-hosting costs and serving limits depend on your deployment.

Access details: DeepSeek open weights (Hugging Face)

The Hugging Face repository. Download the weights and run them yourself or with a host.

deepseek-ai/DeepSeek-V4-Pro-0813

Also on the DeepSeek API as deepseek-v4-pro

Updated Oct 1, 2026

Capabilities & limits

Context window
1M
Maximum output
384K
Image input
Unsupported
Tool calling
Supported
Structured output
Not published

Hosted price

USD / 1M tokens
Input
$1.32
Output
$3.96

What DeepSeek charges for deepseek-v4-pro, its API version of this model.

Free to download. Running the weights costs whatever your hardware or host charges.

Price history
Oct 1, 2026

Input Not applicable per million tokens

Oct 1, 2026

Output Not applicable per million tokens

Technical details

Parameters
1.6T
Active parameters
49B
Architecture
Mixture-of-Experts with hybrid Compressed Sparse Attention and Heavily Compressed Attention, manifold-constrained hyper-connections and the Muon optimizer; speculative decoding module attached
Input modalities
Text
Output modalities
Text

License

MIT

How they describe it

the official release of DeepSeek-V4-Pro, superseding the preview version, with greatly enhanced agentic capabilities

What they say is new

Major Agent upgrades with strong production gains!

Release & lifecycle

Published
Aug 13, 2026
available
As of Oct 1, 2026

Considering a switch?

Specs only go so far. Try both on your own tasks and compare quality, speed and cost.

Compare with Mistral Large 4 →

Prices and availability as of October 20, 2026. DeepSeek's own pages have the latest.