Editorial illustration for Deepseek makes 75% discount permanent, output tokens priced over 34× below GPT‑5.5
Deepseek makes 75% discount permanent, output tokens...
Everyone expected the AI price war to ease up eventually. It hasn’t. Deepseek just dropped the discount hammer for good, making its 75% price cut permanent.
The result is a chasm. Generating a million tokens of text from GPT‑5.5 costs thirty dollars. From Deepseek V4 Pro, it costs less than a dollar.
Thirty-four times cheaper. That’s not a sale. That’s a new reality.
Deepseek V4 clearly trails the top frontier models GPT-5.5 and Opus 4.7 in raw performance. How much exactly depends on the task, and benchmarks only tell half the story; only real-world use will tell. But the price gap is massive, especially for agentic AI systems that chew through many times more tokens than a standard chatbot.
Deepseek isn't playing promotional games. It's removing cost as a factor. A fifty-one times advantage on long outputs shatters the old economics.
Suddenly, that million-token context isn't a premium feature. It's standard. You can afford to generate entire novels worth of text for the price of a coffee.
This forces a brutal question on the rest of the industry. OpenAI, Anthropic, Google. They have to decide if their models are truly fifty times better.
Because if they aren't, their pricing now looks absurd. Deepseek just called their bluff. The response will define the next year.
Related Reading
Business & Startups Nordic pilot adds Gemini for Education, NotebookLM to boost AI literacy
Business & Startups Kling launches Video O1, all-in-one model with MVL bridge using transformer
Business & Startups DeepSeek Seeks More Capital Weeks After USD 7B Funding Round
From Archives Agent explores once, then compiles branch‑free recipe to bypass LLM thereafter
From Archives OpenAI’s Q1 2026 adjusted margin slips to –122%, burning USD 1.22 per USD 1 earned
Common Questions Answered
How much cheaper is Deepseek V4 Pro compared to GPT-5.5 for output tokens?
Deepseek V4 Pro is 34 times cheaper than GPT-5.5 for output token generation. Generating a million tokens from GPT-5.5 costs thirty dollars, while the same output from Deepseek V4 Pro costs less than a dollar, demonstrating a dramatic price difference in the AI market.
What did Deepseek do with its 75% price discount?
Deepseek made its 75% price discount permanent rather than treating it as a temporary promotional offer. This permanent reduction fundamentally changes the economics of AI token generation and removes cost as a competitive factor in the industry.
How does Deepseek's pricing advantage affect long context windows?
With Deepseek's significantly lower pricing, million-token context windows are no longer a premium feature but become standard and affordable. Users can now generate entire novels worth of text for the price of a coffee, making extended context processing economically accessible to everyone.
What challenge does Deepseek's pricing create for competitors like OpenAI and Anthropic?
Competitors must now justify whether their models are truly 50 times better than Deepseek to warrant their higher pricing. If they cannot demonstrate substantially superior performance, their current pricing models become economically indefensible in light of Deepseek's permanent discount strategy.
Further Reading
- DeepSeek cuts V4-Pro prices by 75% — The Next Web
- DeepSeek makes the V4 Pro price discount permanent | Hacker News — Hacker News
- DeepSeek is making its 75 percent API discount permanent — Startup Fortune
- Models & Pricing - DeepSeek API Docs — DeepSeek API Docs
- Tell HN: DeepSeek-V4-Pro 75% discount promo is now permanent — Hacker News
