Table of Contents
CometAI Platform provides powerful controls to customize the behavior of the AI model for each conversation. By adjusting parameters like Temperature and Response Length, you can tailor the output to be more creative or precise, and more concise or detailed, depending on your needs.
- Start a New chat to access the Tuning Parameters for a Conversation.
- Locate the "Model" section at the top of the main chat window.
- Click on the Gear icon to open the Settings panel.
- Once you open Settings, you will see sliders for Temperature and Response Length.
- These controls directly influence the style and length of the AI-generated responses.
The Temperature slider controls the randomness or creativity of the AI's response. You can select a value between 0.0 (Precise) and 1.0 (Creative).
- What is Temperature?
- When using generative AI tools, temperature refers to how precise or how creative you want the output to be. Adjusting the temperature can help you tailor the output to meet your specific needs, affecting how predictable or diverse the response will be.
- Low Temperature (Closer to 0):
- Setting the temperature closer to 0 will produce more deterministic and focused responses. At a low temperature, the model is more likely to select the most probable words or phrases given the context, leading to text that is often more factual and precise but less varied and creative.
- High Temperature (Closer to 1):
- As the temperature approaches 1, the model's responses become more stochastic or random. This means that less probable words or phrases have a higher chance of being selected, resulting in responses that are more diverse, creative, and less predictable.
- When to Use Different Temperatures:
- Setting the temperature of a chat is primarily based on your goal. If you're looking for accurate information, technical explanations, or specific instructions, a lower temperature is generally more suitable. When your goal is to generate creative writing, brainstorm ideas, or explore different perspectives, a higher temperature value can help achieve that by introducing more novelty into the text.
The Response Length setting dictates the verbosity of the model's outputs. A shorter response length yields concise and focused answers quickly, which is beneficial for straightforward questions, rapid insights, and cost reduction, as less computational resources are utilized. Conversely, a longer response length provides more comprehensive and detailed explanations, suitable for complex topics or when a deeper understanding is required. Users should adjust the response length accordingly to balance between succinctness, detail, and potential cost implications of their inquiries.
Important: There is no single "correct" combination, as the ideal parameters depend entirely on your specific task.
The best way to leverage these settings is through experimentation. Try different combinations to see how they affect the AI's output and discover what works best for your unique use cases.
Examples
- To summarize an uploaded document:
- Try a low Temperature (e.g., 0.1) and a Concise Response Length (0).
- To brainstorm marketing slogans:
- Try a high Temperature (e.g., 0.9) and an Average Response Length (3.0).
- To draft a very detailed report section:
- Try a medium Temperature (e.g., 0.5) and a Verbose Response Length (6).
Reasoning Mode controls how much “thinking” the model does before it answers. When enabled, the model can spend extra effort working through a problem, which often improves accuracy on complex tasks but may use more tokens, take longer & cost you more as well.
Which Models Can Use It?
Reasoning Mode is available for Claude and GPT models that support reasoning.
When you select a supported Claude or GPT model, you will see Low / Medium / High / Off near the chat input. If those options are missing, switch to a Claude or GPT reasoning-capable model.
What Each Mode Means
| Mode |
What it does |
Best for |
|
Low
|
Uses an average amount of thinking tokens. Faster replies with lighter reasoning.
|
Simple questions, quick lookups, speed-first tasks
|
|
Medium
|
Adds more reasoning for better accuracy without the heaviest processing time.
|
Everyday questions that need solid accuracy
|
|
High
|
Uses the most thinking effort for deeper analysis. Slower and more token-intensive.
|
Complex problems, multi-step logic, precision-critical work
|
|
Off
|
Disables extended thinking. The model responds without extra reasoning.
|
Fast answers when deep analysis is not needed
|
Default is [Low].
When to Use Which Mode
- Use Low for straightforward prompts, short Q&A, or when you want a quicker reply.
- Use Medium for normal work questions where accuracy matters but full deep reasoning is not required.
- Use High for hard problems: multi-step analysis, careful planning, debugging, or detailed comparisons.
- Use Off when you want the fastest response and do not need extended thinking.
NOTE: Higher reasoning levels can improve thoroughness, but they may increase response time and token usage. Start with Off, Low or Medium, then move to High only when the task needs deeper analysis.