New developments and models in artificial intelligence seem to be coming thick and fast these days. DeepSeek, a language model developed by a team of Chinese researchers and engineers, is making a name for itself in the increasingly competitive field of AI, being touted as a potential rival to ChatGPT.

Like ChatGPT, DeepSeek relies on models trained on huge volumes of text data. Its main strength lies in its ability to understand and generate content in Mandarin and other Asian languages, thus meeting a need often neglected by Western models.

Compared to the American benchmark of OpenAI, DeepSeek stands out for its specialisation in Asian languages, but that's not all. Another distinction is that its latest model (V3) is entirely available as open source, encouraging international collaboration and innovation. This approach enables developers to adapt it to their specific use cases. Finally, DeepSeek boasts a much lower cost than the competition, for more data processed per second.

Among its capabilities, DeepSeek can translate texts and generate content by answering a multitude of questions. It can also code.

While intelligent conversational agents are popping up everywhere (like ChatGPT, Gemini, Claude, Mistral to name only the most famous), DeepSeek presents itself as an original alternative, which can be tested now: deepseek.com.

– AFP Relaxnews