Artificial Intelligence is very far from being a new concept. I mean, look back at 2001, and there was a Spielberg movie named after it. However, the modern form of “AI” really only blew up in late 2022 with the release of OpenAI’s ChatGPT. That one product spawned the rapid release of several competitors like Gemini and Copilot.
Now, there’s a new competitor in the ring, and despite the fact that it is coming from a company that almost all of us have never heard about, it might be the biggest disruptor in the space in a very long time. It’s called DeepSeek and it seems that the public response to it should have other AI platforms very worried.
Chinese Start-Up DeepSeek Has Its AI Model Rocket to the Top of App Store Charts
The AI market has seen many new competitors in the last two years or so. It all started with ChatGPT, and then options like Copilot, Claude, and Google Gemini eventually revealed themselves. However, through it all, ChatGPT has remained king, with its name almost synonymous with any AI chatbot. That’s a pretty enviable position to be in and it seemed quite unshakeable.
However, it seems that things have been shaken up, and by quite a lot too. This is where China comes in. An AI-focused start-up out of the Asian country, named DeepSeek, recently released its DeepSeek-R1 model. However, it was clear that things were very different when the model was able to climb straight up Apple’s US app charts, soaring past ChatGPT’s position to the #1 spot. That’s notable.
And here’s the thing, it is easy to understand the things that make DeepSeek so attractive. For one, it is open source, which means that the codebase is available and open for scrutiny, improvements, and modifications. That’s not something OpenAI’s ChatGPT will be emulating. Plus, it is deemed to be a lot more efficient, running on chips that are a lot less powerful than usual for AI applications. It’s definitely standing out on its own merit.
DeepSeek Is Very Likely to See Rapid Adoption, Particularly Due to Its Low Costs
A lot of software platforms that utilize AI in some capacity usually turn to OpenAI to power things. However, using these AI models isn’t free. For every one million input tokens in OpenAI’s o1 model, you pay $15, and for every one million output tokens, that becomes $60. For a big company, that might not be a lot of money, but if you’re maximizing profits, a lower-cost solution never hurts.
That’s exactly where DeepSeek is likely to come in. The DeepSeek-R1 model costs only $0.55 for every million input tokens and $2.19 for a million output tokens. Comparatively, that’s insane. And when you consider that the hardware needed to make the model run doesn’t need to be as high-powered, and consequently, cheaper, it seems like a pretty big win for the Chinese start-up.
For a few million dollars, a Chinese start-up (with limited access to higher-end chips) has done what US companies utilize billions for, so it’s not hard to see why it’s a massive worry.
