DeepSeek in Bubble: Let the discussions begin

recker249 · 2025-01-28T04:48:43.912Z

Is anyone already looking at incorporating Deepseek into their Bubble apps?

mitchbaylis · 2025-01-28T06:27:49.972Z

I swapped out the chatgpt apis to deepseek for a few apps - it’s working very well

the o1 wasn’t available on all my chatgpt accounts so it was a pretty simple decision since I needed o1 capabilities

pretty wild what it can do for a fraction of the cost of chatgpt

mdwp.post · 2025-01-28T14:40:42.918Z

Two of my templates use Deepseek.

jonathan.baxter1991 · 2025-01-28T23:08:57.331Z

Nice, is the setup much different from openai? Are you using the API connector?

janernestgo · 2025-01-29T23:45:18.913Z

Is it helpful (smarter than gpt) in guiding bubble.io development?

ihsanzainal84 · 2025-01-30T04:16:33.321Z

With similar capabilities from these great models, the only difference that matters, is price. It’s way cheaper to run Deepseek V3 or R1, even on pay as you go inference providers like Azure and Cloudflare. The point is getting the required results.

After tons of research, I realized that I too got swept up by the AI hype. When I narrowed down the use cases of what I want to achieve with AI, I found cheaper, more specialized models that can do what I need at lower costs and faster speeds (distilled models are naturally faster due to their size).

While it’s great to be able to use the latest and the greatest, if all your’re doing is just summarizing some text and return a JSON, you can find smaller models to run in the cloud that will achieve the same results. Just do some testing, fine-tuning and you’re good to go. Just use AI to refine your AI.

mdwp.post · 2025-01-31T12:28:18.648Z

If you use DeepSeek in the Groq cloud your data will not be send to servers in China. GroqCloud™ Makes DeepSeek R1 Distill Llama 70B Available - Groq is Fast AI Inference

Note that Groq is a US-based company and when you send a query via DeepSeek-R1-Distill-Llama-70b on GroqCloud, we’ve deployed it on our infrastructure. In particular, when you send a query (inputs/outputs) to any model instance via GroqCloud™, we temporarily store it in memory and clear it as soon as your session is completed. We do this because we don’t do training – we do inference. While Groq has no incentive to keep your data, customers who want storage can work with their own provider to meet their application needs.

This means your data won’t be sent to any DeepSeek servers in China. We take privacy seriously at Groq and you can learn more about that commitment at trust.groq.com.