This is what DeepSeek China does better than ChatGBT
 |
| DeepSeek |
A Chinese startup has launched a new artificial intelligence model called DeepSeek, and it's generating a lot of buzz in the US tech sector.
Why? Because it outperforms all other major companies' models. Even more impressively, it claims to have achieved this with less funding and technical resources.
Sky News published a report today comparing DeepSeek's performance to OpenAI's popular chatbot, ChatGPT.
When it comes to AI output, opinions can vary depending on the use case. Therefore, there are still areas where other models might outperform DeepSeek.
For example, some users have discovered that DeepSeek's chatbot responses are sometimes censored due to Chinese government policies. This is true.
But here's the thing: DeepSeek is open in a way that OpenAI promised with ChatGPT but failed to deliver. In fact, while OpenAI abandoned its initial commitment to "openness," DeepSeek launched its model as open source. Anyone can download the DeepSeek R1 model for free and run it locally on their own devices.
This means your data will not be shared with DeepSeek in any way. Furthermore, as the company itself points out, users can bypass any censorship or biased results.
Read more : Snapdragon Wear Elite arrives with on-device AI capabilities, performance and battery gains
While training a single OpenAI model can cost tens of millions of dollars, DeepSeek claims to have trained its model for only $5.5 million. This cost difference is also reflected in the user experience.
Access to the DeepSeek R1 model's API starts at $0.14 per million characters (approximately 750,000 words). In contrast, a similar OpenAI model (o1) is priced at around $7.50 per million characters. That's a significant price difference.
Users of both ChatGPT and DeepSeek agree that the OpenAI bot still excels in creative or conversational output, as well as in providing information related to current events.
But the general consensus is that DeepSeek excels at technical tasks. If you use chatbots for logical reasoning, programming, or solving mathematical equations, you may find DeepSeek's output superior.
For most other queries, however, the two models appear to deliver roughly equal results.
Even this parity is bad news for OpenAI and ChatGPT because DeepSeek is completely free for most use cases, while regular ChatGPT users need to subscribe to the paid plan for $20 per month.
For companies that use APIs for AI models, the price difference between two nearly equivalent models may be significant enough to warrant switching from ChatGPT to DeepSeek.
Read more : Australia bans AI services for teenagers
