DeepSeek V4-Flash Impresses: 40 Seconds vs GPT's 1:47 on Same Task, Cost Tens to Hundreds of Times Lower
DeepSeek
OpenAI
DeepSeek's V4-Flash model completes the same task in 40 seconds, while the most powerful GPT model takes 1 minute 47 seconds. Additionally, DeepSeek's pricing is tens to hundreds of times lower than GPT's.
A news report highlights the impressive performance of DeepSeek V4-Flash, the official version. In a benchmark, the model completed the same task in just 40 seconds, whereas the most powerful model from GPT took 1 minute and 47 seconds. Furthermore, the cost of using GPT is said to be dozens to hundreds of times higher than that of DeepSeek, emphasizing the efficiency and affordability of DeepSeek V4-Flash.
- Сокращения
- GPT = Generative Pre-trained Transformer — генеративный предобученный трансформер
Source: DeepSeek (GNews) —
original