Overview
AI models are sophisticated algorithms trained on vast amounts of data to perform various tasks, from text generation to analysis and problem-solving.
They are the AI engines that power ModularMind’s capabilities. Understanding and selecting the right model for your task helps ensure optimal performance and cost-efficiency.
Usage
In ModularMind, each stack is powered by a model of your choice. You can select a model based on the complexity or the type of task you want to perform. For example, for simple data extraction task, you can use models like GPT 3.5 or Claude 3 Haiku, while for harder tasks like creative copy writing and document intelligence, you can use models like GPT 4 Turbo or Claude 3.5 Sonnet.
You can pick the model for a stack by using the dropdown menu above the stack on the board, or click on the stack to open the stack details and select the model from the dropdown menu on top of the stack details panel.
When you run the stack, the prompts inside it will be merged and sent to the model of your choice for processing to generate the output. You can also change the model for a stack at any time by selecting a different one from the dropdown menu.
Model Details
Each model has its own unique capabilities and limitations. You can find brief descriptions of each model in the dropdown menu next to the model name.
To learn more about a specific model, open settings from top menu and click on the Models tab. Here you can find detailed information about each model, including its provider, capabilities, and cost.
Credits
ModularMind uses a credit system for you to easily access state-of-the-art AI models, without worrying about API keys and service access to providers. You can see your credit balance by clicking on the Credits button in the top menu. You can see how much tokens you can process with your current credit balance by hovering over the credit balance.
Here you can also purchase more credits if needed, or invite colleagues to earn free credits. You will be rewarded with free credits when your invitees sign up and make their first purchase with your invitation link.
In the Models tab in settings, you can see the cost of each model in credits per token by hovering over the model cost.
You can test different models to see which one works best for your task and optimize your workflow by selecting the most cost-effective model for your needs.
One rule of thumb is to select the more costly models for complex tasks where accuracy and sophistication are paramount, and opt for more cost-effective models for straightforward tasks like data extraction.
Using API Keys
If you have your own API keys for a specific model, you can use them in ModularMind by clicking on the Use Your API Key button next to the name of the provider. This will open a dialog where you can enter your API key.
Currently ModularMind supports API keys for OpenAI models.
Remember to experiment with different models to find the best fit for your specific use case. ModularMind’s flexible interface allows you to easily switch between models, enabling you to optimize your workflow for both performance and cost.