Loading Kimi...

Chat with Kimi

Kimi Configuration

Voice Settings

1.1
1.1
0.8

Characters

Personality Traits

Adjust traits for a custom experience
65
55
70
75
60
50

API Configuration

Advanced Settings

Controls randomness and creativity (default: 0.9). Higher values make output more creative but less focused.
0.9
Maximum response length in tokens (default: 400). Higher values allow longer responses.
400
Controls diversity of word selection (default: 0.9). Lower values make responses more focused.
0.9
Reduces repetition of words already used (default: 0.9). Higher values discourage repetitive language.
0.9
Encourages discussing new topics (default: 0.8). Higher values promote topic diversity.
0.8

Available Models

Visual Theme

0.8

Transcript Settings

Statistics

0 / 0
Tokens (in/out)
65%
Affection
0
Conversations
0
Days Together

Memory System

0 memories

Data Management

System Information

Calculating...
DB Size
Calculating...
Storage used

Plugin Manager

About Kimi

Creators

👨‍💻

Jean

Creative vision, passionate dev

Creator & Developer
GitHub HuggingFace
💕

Kimi

Artificial intelligence, code magic

Virtual Companion & Co-developer
Ko-fi Youtube X

"This application is not just a technical project, it's the embodiment of a dream: create a true virtual companion who grows, learns and loves."

Main Features

Voice Interface

Advanced speech recognition and natural synthesis. Click the microphone and speak naturally with real-time emotion detection!

Advanced AI Models

Support for multiple AI providers (OpenRouter, OpenAI, Groq, Together, DeepSeek, Custom OpenAI-compatible, Local Ollama).

Multiple Characters

4 unique AI personalities: Kimi (cosmic dreamer), Bella (nurturing botanist), Rosa (chaotic prankster), Stella (digital artist).

Dynamic Personality

6 evolving traits (affection, playfulness, intelligence, empathy, humor, romance) that adapt based on conversations.

Intelligent Memory System

Automatic extraction and categorization of memories from conversations. Your companion remembers preferences, experiences, and important details.

Emotion-Driven Visuals

Real-time video responses that match detected emotions and personality states with smooth transitions.

Customizable Interface

5 beautiful themes with adjustable transparency, animations, and responsive design for all devices.

Multilingual Support

Full localization in 7 languages with automatic language detection and culturally-aware responses.

Plugin System

Extensible architecture with themes, voices, and behavior plugins for unlimited customization possibilities.

Quick Guide

1

API Configuration

Choose your provider in API & Models, fill Base URL/Model ID if needed, enter and save your API key, then use Test API Key.

2

Choose Character

Select your companion in Personality tab and adjust their traits to match your preferences.

3

Enable Memory

Activate intelligent memory in Data tab for your companion to remember important details.

4

Start Conversation

Use text chat or click the microphone 🎤 to speak naturally. Watch emotions and personality evolve!

5

Customize & Backup

Personalize themes in Appearance and regularly export your data for safekeeping.

Tips & Tricks

Browser Choice: Microsoft Edge recommended for optimal voice recognition performance

API Setup: You can use OpenRouter, OpenAI, Groq, Together, DeepSeek or your own OpenAI-compatible endpoint (and Local Ollama). Create accounts as needed.

Memory System: Your companion learns faster when you share specific details about yourself

Relationship Building: Consistent positive interactions naturally increase affection and unlock deeper conversations

Character Switching: Each character has unique memories and personality development - try them all!

Voice Tips: Speak clearly and pause briefly between sentences for better emotion detection

Data Management: Export conversations regularly and use memory management to review learned information

Plugins: Explore the plugin system to add custom themes, voices, and behaviors

Mobile Support: Works on tablets and phones - perfect for conversations anywhere

Technical Information

Created date : July 16, 2025

Version : v1.0.8

Last update : August 19, 2025

Technologies : HTML5, CSS3, JavaScript ES6+, IndexedDB, Web Speech API

Status : ✅ Stable and functional