JADE

JADE

Cheap Multi-Models API chatbot

8 followers

JADE is a chatbot to use and compare different LLMs while saving money. You can ask the same questions to different LLMs, pick the best answer and continue the conversion with it. So answer from one model can be use by another and pay just for what you use.
JADE gallery image
JADE gallery image
JADE gallery image
JADE gallery image
JADE gallery image
Free Options
Launch Team / Built With
Unblocked AI Code Review
Unblocked AI Code Review
High-signal comments based on your team's context
Promoted

What do you think? …

Adrien Bouvais
I'm thrilled to announce the launch of JADE, a chatbot app that lets you use and compare multiple Large Language Models (LLMs) while saving money. As a frequent user of LLMs, I found myself asking the same questions to different models, wanting to compare their answers and pick the best one. But with multiple subscriptions and limited compatibility, it was a costly and frustrating experience. That's why I created JADE - to make it easy and affordable for anyone to access multiple LLMs within a single conversation. What sets JADE apart: - Use multiple models from 10+ providers, including OpenAI, Anthropic, Google, and more - Compare responses and choose the best one - Use the selected response across all models for seamless conversation - Save money with API pricing and pay only for what you use. For example, I pay just $3/month for around 1500 messages across Gemini, Llama, Claude, and GPT-4! Perfect for: - Developers seeking the best coding solutions - Writers looking for the perfect phrase - ChatGPT Plus users who want to save money and explore alternative models - Anyone wanting to benchmark different LLMs
Ryden Sun
Congrats on the launch, @mrbounty! JADE sounds like an amazing tool for efficiently comparing LLM responses. The ability to ask multiple models and select the best answer is a game-changer. Love that it’s budget-friendly too! Can’t wait to try it out and see how it works in practice. Upvoted! 🚀
dashuai
This sounds super interesting! How do you ensure the compatibility between different LLMs when using JADE? Will there be any limitations on the number of questions per model?
Adrien Bouvais
@ezshine Thanks! API do not really care about where the text come from, you just give it a list of text (messages) and it give you back one more. So when you ask something, it retrieve the same list of messages that is saved in my database, turn it into the right JSON format and do a curl request to each provider server. There isn't any limitations other than the context lenght. As they all use different tokenizer, I currently do not check if each message is too long for a corresponding model. So you can find an error like "Message too long".