00:00 Speaker A
Jeff, what has implementation been like for your company and what do companies actually do when it comes to implementation?
00:06 jeff
As you know, Olama continues to add nearly 1 million new installs every week. From there, it was about how to take it to a new level by using Olama’s cloud to deliver these new frontier-level open models. Correspondingly, token usage has doubled on average every month for the past six months. Therefore, the demand for open models is truly insatiable.
00:35 Speaker A
Is there a demand for this project, or a Fortune 500 company?
00:41 jeff
absolutely. And a little-known fact is that the majority of Fortune 500 companies already use open models. It is used everywhere from retail to the most regulated industries such as healthcare and industrial businesses. So what’s really important to them is, how can I bring this AI closer to my data and actually extend their existing workflows without risking setting it up in a third-party model provider or another region where they can’t do that because of business compliance requirements?
01:21 Speaker A
Jeff, what is your business model? How do you all make money?
01:27 jeff
Olama has subscriptions that individual developers can start. It’s free to download, and if you need more power with the largest open model hosted in the cloud, Olama offers subscriptions to access UH servers in the US and Europe. Um, we’re really excited about the introduction of the Team Plan for businesses and enterprises that will enable them to adopt an open model across their business. So it will be very similar. You know, starting with a subscription fee and being able to predictably scale the price of your tokens, you can actually invest in massive amounts of compute and AI to power any kind of business.
02:22 Speaker A
I often hear Jeff talk about the open source model in China. After all, is it a long-term competitive threat?
02:32 jeff
I think when we talk to customers, what’s more important to them is that they think about where the model came from. More importantly, however, is where and how the model is run, and that it can be run in the safest way. Well, we’re seeing incredible growth for Nvidia’s new Nemotron 3 Ultra model. These are the best performing models for long-running agent tasks. We’re really excited about these things, and so are our customers.
