by DeepSeek· Released May 2025· Cutoff May 2025
DeepSeek-V4-Flash is DeepSeek's primary flagship model, optimized for fast and cost-effective inference. It offers a massive 1M token context window and supports multimodal inputs including images and audio. As the latest iteration, it balances high performance with significantly lower pricing compared to its predecessor.
Input cost
$0.05 per 1M tokens
Output cost
$0.25 per 1M tokens
Context window
1M tokens
Max output
8192 tokens
Modalities
License
proprietary
High-speed, cost-sensitive applications requiring large context and multimodal understanding.
