DeepSeek Slashes Off-Peak Prices by Up to 75% to Balance AI Demand

DeepSeek Introduces Steep Off-Peak Discounts

Chinese AI startup DeepSeek has announced significant price reductions for its API services during off-peak hours, aiming to balance out the surging demand for its artificial intelligence models. Starting Wednesday, the company is offering discounts of up to 75% for accessing its application programming interface (API) during specific timeframes 1

1

Pricing Strategy and Model Access

The pricing strategy affects two of DeepSeek's primary models:

1. DeepSeek V3 model: 50% discount from 12:30 a.m. to 8:30 a.m. Beijing time.
2. R1 model (powering DeepSeek's chatbot): 75% price reduction during the same period 2

   2

   .

Specifically, during off-peak hours (16:30 - 00:30 GMT), the costs for DeepSeek-V3 are reduced to $0.035 per million tokens for input (cache hit), $0.135 for input (cache miss), and $0.550 for output. The DeepSeek-R1 model sees even steeper discounts, with prices dropping to the same levels as V3 5

5

Managing High Demand and Competition

This move comes as DeepSeek grapples with overwhelming interest from users and developers adopting its technology. The company recently reopened top-ups for API access credits, though it cautioned about constrained capacity during peak hours 1

1

DeepSeek's technology has gained traction among major players, including:

• Tencent Holdings Ltd.
• Perplexity AI
• Various Chinese state entities, including Hong Kong government departments 3

  3

  .

Related Stories

DeepSeek raises API pricing up to 1,100%, reopens $8bn round, invests in robotics

DeepSeek Limits API Access Amid Surging Demand for Its AI Model

DeepSeek V4 Pro launches with fourfold price increase as AI demand strains infrastructure capacity

DeepSeek Resumes API Access Amid Intensifying Chinese AI Competition

Future Developments and Industry Impact

DeepSeek is not resting on its laurels. The company has announced plans to release its next reasoning model, DeepSeek R2, earlier than the initially planned May launch. This new model is expected to offer improved coding capabilities and reasoning in multiple languages beyond English 5

5

The startup's approach to AI development has challenged industry norms:

1. Cost-effective innovation: DeepSeek's models offer performance comparable to OpenAI at a fraction of the cost 3

   3

   .
2. Open-source commitment: The company launched an "open source week," releasing five new repositories, including DeepGEMM, an FP8 GEMM library optimized for dense and Mixture of Experts (MoE) computations 5

   5

   .

Market Implications and Future Outlook

DeepSeek's pricing strategy and technological advancements have significant implications for the AI industry:

1. Increased competition in the Chinese AI ecosystem, with companies like Alibaba introducing new models and investing heavily in AI infrastructure 5

   5

   .
2. Potential shift in the global AI landscape, challenging established players like OpenAI on both performance and cost 3

   3

   .
3. Growing interest from investors, with DeepSeek considering outside funding to support its expansion and infrastructure needs 3

   3

   .