Explorer Tier: Investing in AI entrepreneurs
Explorer Tier: Investing in AI entrepreneurs
At Nebius, we’re committed to supporting the AI community, which is why we’re introducing our Explorer Tier, designed to make GPUs more affordable and accessible for those just getting started. H100 is now priced at $1.5/h for the first 1,000 hours monthly.
Nebius is committed to supporting and investing in the AI entrepreneur community. We sponsor hackathons and interact with the community of startups, both big and small, on a daily basis. With up to 70% of AI funding being spent on compute, we still recognize the challenges that high GPU costs pose for getting a project off the ground.
We also see the amazing things that hackers can accomplish in a short time, often with as few as 1–2 H100 GPUs. However, securing access to GPUs can be challenging, with limited availability of H100 in some regions and the requirement to use full hosts with 8 GPUs. Some clients report needing to keep their instances running 24/7 to avoid stockouts, making on-demand usage difficult, if not impossible, in practice.
What we offer
Simply sign up and use up to 8 GPUs at a time, whenever you need them. The first 1,000 H100 GPU-hours would cost you only 1.50 USD per hour. We offer the lowest price on the market, with genuine self-service. No commitments. No stockouts.
The Explorer Tier is valid for all new customers who join after October 1, 2024. These prices are valid until the end of 2024, and the offer may be extended into 2025 or adjusted based on customer demand and feedback.
How it works
Let’s take a look at the table below, illustrating in detail how these conditions work.
For example, if you deploy one VM with one NVIDIA H100 GPU at 00:00 on November 1 and run it the entire month (720 hours), you’re within the first 1,000 hours of the Explorer Tier, so the price is 1.5 USD per GPU/hour. This price includes CPU and RAM. The total cost at the end of the month will be 1,080 USD.
If you deploy two VMs under the same conditions, your total GPU operating time will be 1,440 hours. The Explorer Tier price applies to the first 1,000 hours, and the remaining 440 hours will be charged at the standard on-demand price of 2.95 USD per hour. In this case, the total cost will be 2,798 USD (1,500 USD + 1,298 USD).
Please keep in mind that all storage will be billed separately.
Why are we doing this?
At Nebius, we firmly believe that our long-term success is intertwined with the success of the AI entrepreneur community. It’s in our best interest to support those bold enough to try to change the world — because the’re the ones who do.
We want to be part of your success. By offering market-leading pricing when you’re just getting started, we aim to ease your path to a working product, funding and revenue. In the meantime, we hope you’ll experience our full-stack platform and make it your weapon of choice by the time you’re ready to scale.
See also
Our Paris availability zone will be one of the first in Europe to offer NVIDIA H200 Tensor Core GPUs. This is just the beginning as we continue to expand our operational capacity to meet fast-growing demand for high-quality infrastructure and the latest GPUs from AI builders around the world.
We have developed a new version of Nebius platform that we believe will serve your needs even better. It is already being tested by our in-house LLM R&D team and a number of clients. Now, we are rolling it out to everyone.