CHAET - C.H.E.T.

Chat Helper for (almost) Every Task

Model & Settings

Model Parameters

Max Tokens: 1024

Maximum tokens to generate in the response.
Higher values allow longer responses but cost more. Each model has different limits.

Temperature: 0.6

Controls creativity and randomness.
0 = Deterministic, focused responses
1 = Balanced creativity
2+ = More creative, less predictable

Top P: 0.9

Nucleus sampling threshold.
Controls diversity by limiting token choices to cumulative probability mass. Lower = more focused, higher = more diverse.

Top K: 40

Vocabulary filtering.
Limits AI to choose from top K most probable words. Lower = more focused, higher = more variety.

Repetition Penalty: 1.1

Prevents repetitive text.
1.0 = No penalty, >1.0 = Discourage repetition, <1.0 = Allow repetition.

Frequency Penalty: 0.0

Reduces likelihood of repeating exact phrases.
Positive values discourage repetition, negative values encourage it.

Presence Penalty: 0.0

Encourages introducing new topics.
Positive values promote new concepts, negative values focus on current topics.

Seed (Optional):

Reproducibility control.
Set a number for consistent outputs with same inputs. Leave empty for random results.

Welcome to C.H.E.T.

Add Prompt

Add MCP Server

Server Response Metadata