Settings
API Key
The API key setting is used to securely authenticate and authorize access to our API. An entry in this field is required for the software to function. Please contact PrimisAI support in case you have not yet been provided with an API key.
Username
This setting allows you to provide a custom username to be displayed in the chat interface. You can enter your preferred name or alias in this field to personalize your interaction with RapidGPT. If left empty, the assistant will automatically use the full name obtained from the authentication mechanism.
Precision
Adjusts the AI's precision to suit different preferences. The mood setting ranges from "More Precise" for accuracy-focused outputs, to "Balanced" (the default) for a well-rounded approach, and to "More Creative" for outputs that prioritize creativity and innovation.
Interactivity Rate
This setting controls how much RapidGPT engages with you in response to your prompts. A higher interactivity rate means that the AI will ask more follow-up questions and provide more detailed responses. A lower interactivity rate means that the AI will be more concise and direct in its responses.
Enable RAG
This setting allows Pro
or Enterprise
users to enable or disable RAG (Retrieval-Augmented Generation) in their conversations with RapidGPT. RAG is an advanced feature that enhances the AI’s responses by retrieving relevant information from a knowledge base before generating an answer:
-
When enabled, RapidGPT will utilize RAG to provide more informed and context-aware responses, leveraging a dedicated repository of information to supplement its knowledge.
-
When disabled, RapidGPT will rely solely on its pre-trained knowledge to generate responses. Please note that this setting has no effect for
Free
tier users, as RAG functionality is exclusively available toPro
andEnterprise
users. If you're aFree
tier user and wish to access this feature, consider upgrading your subscription.
Pro
and Enterprise
users can toggle this setting on or off based on their specific needs for each conversation, allowing for flexibility in how they interact with RapidGPT.
Personal System Prompt
This setting allows users to provide custom instructions to the assistant, enabling a more tailored interaction experience. By utilizing this feature, you can shape the AI's behavior to align with specific requirements, preferences, or standards in your work environment. The personal system prompt can be used to customize various aspects of the AI's responses, including:
- Coding standards and practices
- Language preferences
- Terminology and verbiage
- Output format and style
- Specific domain knowledge focus
For example, you might use the personal system prompt to instruct the AI to:
- “Use logic instead of wire/reg in SystemVerilog files.”
- “Always reply in Japanese.”
- “Provide explanations suitable for junior developers.”
- “Include performance considerations in all discussions revolving around design.”
- or a combination of the above.
By setting a personal system prompt, you can ensure that RapidGPT responses are consistently aligned with your project needs, team standards, or personal preferences, enhancing the efficiency and relevance of your interactions with the AI assistant.
Remote Server URL
By configuring the Remote Server URL setting, you can specify the URL path and HTTP method for an alternative endpoint. This allows you to define your own API route that corresponds to a specific RapidGPT resource within your organization.
Message Send Mode
This setting allows you to customize the keystroke used for sending messages in the chat interface. You can choose between two options:
-
Enter
Only: Select this option if you prefer to send messages by pressing theEnter
key alone. This setting is ideal for users who value speed and efficiency in their interactions. -
Ctrl
+Enter
: Choose this option if you want to send messages by pressing theCtrl
key along with theEnter
key. This setting is useful for those who want to avoid accidentally sending incomplete messages and prefer having a confirmation step before sending.
Chat History Path
Specify an optional path to save the chat history. By default, RapidGPT manages the chat history project-wise. However, using this option allows you to override the default behavior and save the chat history to a specific location. Please note that enabling this option will supersede the project-wise chat history management.
AutoReview Max Iterations
This setting enables users to define a custom iteration limit within AutoReview, effectively reducing the incidence of 429
errors. While increasing the iteration count can enhance the likelihood of RapidGPT finding a solution, it is important to note that execution times may extend proportionally.