DeepSeek V4 Pro has 1.6T total parameters, its largest model by that metric, and V4 Flash has 284B parameters; both models have a context window of 1M tokens

· Source ↗

TLDR

  • DeepSeek released V4 Pro (1.6T parameters) and V4 Flash (284B parameters), each with a 1M token context window.

Key Facts

  • V4 Pro is DeepSeek’s largest model by parameter count at 1.6 trillion total parameters.
  • V4 Flash is the smaller variant with 284 billion parameters.
  • Both models support a 1 million token context window.
  • DeepSeek claims V4 is competitive with top closed-source models from OpenAI and Google DeepMind.

Why It Matters

  • V4 Pro’s 1.6T parameter count marks a significant scale increase for DeepSeek relative to its prior releases.
  • The competitive claim against leading closed-source models, if substantiated, would extend DeepSeek’s pattern of narrowing that gap.

Vincent Chow / South China Morning Post · 2026-04-24 · Read the original