Skip to content
AI Model Radar

DeepSeek V4 Flash 0731

deepseek-ai/DeepSeek-V4-Flash-0731
Open Source · chat · open-weights
Open Alert me on changes
Context
1M
Max output
384K
Weights
Open
API $/1M
$0.14 / $0.28
Modalities
text
Released
31 Jul 2026
License: mit · deepseek-ai/DeepSeek-V4-Flash-0731
Download image Share on X Share on LinkedIn
AI summary
● machine-written

DeepSeek V4 Flash 0731: Sparse MoE model with 13B active parameters

DeepSeek V4 Flash 0731 is a sparse mixture-of-experts text generation model with 13B active parameters out of 284B total, supporting a 1M-token context window. The re-post-trained revision is suited for coding, reasoning, and agent workflows. It is available under the MIT license as an open-weights model.

What's new
  • 1,048,576 token context window (1M tokens)
  • 13B active parameters out of 284B total sparse MoE architecture
  • 384,000 token maximum output capacity
  • Released July 31, 2026
Best for
Coding assistantsReasoning tasksAgent workflowsChat systems
Sources

Source: https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731