DeepSeek claims 545% profit margin for AI models based on theoretical income

Chinese AI startup DeepSeek has announced that its AI models could be highly profitable, albeit with caveats, revealing a theoretical profit margin of 545% for its online services.

In a post on X, DeepSeek outlined that this margin was calculated based on “theoretical income” rather than current revenue. The company elaborated on these figures in a GitHub post, explaining its strategy for “higher throughput and lower latency.” According to DeepSeek, a 24-hour period of usage for its V3 and R1 models would generate $562,027 in daily revenue if billed entirely at R1 prices. The operational cost of leasing the required GPUs (graphics processing units) was estimated at $87,072.

🚀 Day 6 of #OpenSourceWeek: One More Thing – DeepSeek-V3/R1 Inference System Overview

Optimized throughput and latency via:
🔧 Cross-node EP-powered batch scaling
🔄 Computation-communication overlap
⚖️ Load balancing

Statistics of DeepSeek's Online Service:
⚡ 73.7k/14.8k…

— DeepSeek (@deepseek_ai) March 1, 2025

However, DeepSeek acknowledged that its actual revenue is 'substantially lower' due to several factors. These include nighttime discounts, lower pricing for the V3 model, and the fact that only a portion of its services are monetised, with web and app access remaining free. The company also recognised that removing free access and discounts would likely reduce usage, making the projected profit margins speculative rather than reflective of its current financial standing.

The disclosure comes amid wider discussions about the cost and profitability of AI technology. DeepSeek gained attention in January after unveiling a model that reportedly matched OpenAI’s o1 on certain benchmarks, despite being developed at a lower cost and without access to the most advanced chips due to US trade restrictions.

DeepSeek’s technology has not only drawn the interest of industry observers but also made a brief impact in the consumer space. Its app momentarily surpassed OpenAI’s ChatGPT at the top of Apple’s App Store rankings, though it has since dropped off the general rankings and currently stands at sixth place in the productivity category, behind ChatGPT, Grok, and Google Gemini.