I'm looking for some assistance in estimating the budget for deploying a private version of DeepSeek, specifically the R1 and V3 models. This setup needs to support a company with around 3500 employees. Any insights or guidance would be greatly appreciated!
4 Answers
Also, don’t forget to factor in hardware costs! Investing in a couple of Mac Studios with 512GB RAM might be necessary too, depending on your usage.
In my experience from working at OpenRouter, DeepInfra and Novita have been solid choices for such projects. For a company that size, I’d recommend connecting directly to the DeepSeek API for better integration.
You might want to reach out to DeepInfra, Modal, or any AI solutions provider. They usually have options for private deployments that could fit your needs.
Thanks for the suggestions!
I found a great article that might help you out—check it out for some additional insights on cost estimation for local deployment: https://www.aifun.cc/en/deepseek-local-deployment-cost.html
Just a heads up, OP mentioned they need Private inference!