Deepseek makes its 75 percent discount permanent, pricing output tokens at least 34x below GPT-5.5

Editor's take

Deepseek has solidified its aggressive pricing strategy by making a significant discount on its V4-Pro model a permanent fixture, positioning its output tokens at a fraction of what competitors like OpenAI's anticipated GPT-5.5 might charge.

This move directly targets developers building agentic systems or applications requiring high-volume token processing, where cost is a primary constraint. By undercutting leading models by such a substantial margin, Deepseek positions itself as a compelling alternative for use cases previously deemed too expensive, potentially accelerating the adoption of more complex AI-driven workflows.

Future developments will reveal if other major players like OpenAI or Anthropic will respond with similar price adjustments for their flagship models, or if Deepseek's strategy will force a broader re-evaluation of LLM token economics. The long-term impact on market share and the viability of cost-sensitive AI applications hinges on this competitive response.