Skip to content
AI Model Radar

DeepSeek V4 Pro 0813

deepseek-ai/DeepSeek-V4-Pro-0813
Open Source · chat · open-weights
Open Alert me on changes
Context
1M
Max output
384K
Weights
Open
API $/1M
$0.43 / $0.87
Modalities
text
Released
13 Aug 2026
License: mit · deepseek-ai/DeepSeek-V4-Pro-0813
Download image Share on X Share on LinkedIn
AI summary
● machine-written

DeepSeek V4 Pro 0813: GA release of mixture-of-experts model

DeepSeek V4 Pro 0813 is the general availability release of a large-scale mixture-of-experts model with 1.6T total parameters and 49B activated parameters. It supports a 1M-token context window and maximum output of 384,000 tokens. The model is designed for advanced reasoning, coding, and long-horizon agent workflows.

What's new
  • GA release of DeepSeek V4 Pro line
  • 1M-token context window with 384,000 token max output
  • Hybrid attention system for efficient long-context processing
  • Supports reasoning efforts high and xhigh
Best for
Advanced reasoning and coding tasksFull-codebase analysisMulti-step automation and long-horizon agent workflowsLarge-scale information synthesis
Sources

Source: https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro-0813