Multi LLM AI Chat by fatalv.nu

Multi-AI Chat Assistant

A powerful browser extension that brings multiple AI models into one unified chat interface — chat, compare, generate images, and more, all from your browser sidebar.

Features

Chat

• 31+ AI providers — OpenAI, Anthropic Claude, Google Gemini, DeepSeek, Grok (xAI), Groq, Mistral, Cohere, Perplexity, Hugging Face, Ollama (local & cloud), OpenRouter, Together AI, and many more
• Streaming responses — See AI replies in real-time with typing animation
• Compare mode — Run prompts against two models simultaneously and compare answers side-by-side
• Thinking/reasoning display — View model reasoning in expandable sections (DeepSeek-R1, Qwen, Ollama, etc.)
• System prompt — Set custom system instructions per conversation
• Context window control — Override max output tokens per request
• Temperature & Top-P — Fine-tune generation parameters with sliders
• Edit messages — Click any user message to reload it for editing and re-sending
• Retry last prompt — Quick re-send of the previous message

Image Generation

• Multiple image APIs — Google Gemini (Imagen 4, Nano Banana), Stable Diffusion, Grok Imagine, DeepInfra FLUX, SiliconFlow, Together AI, Novita AI
• Quality & resolution controls — Select quality (Standard/HD), aspect ratio, and resolution
• Multi-image generation — Generate 1–4 images at once
• Image lightbox — Click images to enlarge, copy, or download as PNG/JPG

Conversations

• Multi-chat management — Create, rename, favorite, pin, and delete conversations
• Search & filter — Find conversations by title instantly
• Multi-select — Batch delete or export multiple conversations at once
• Conversation branching — Fork from any AI response to explore alternative paths
• Import/Export — Backup and restore chats in JSON, Markdown, or plain text
• Export to PDF — Save conversations as print-friendly documents
• Prompt templates — Save and reuse frequently used prompts and system instructions

Sidebar & UI

• Persistent sidebar — Pin the chat as a browser sidebar that follows you across tabs
• Collapsible panel — Toggle sidebar visibility with one click
• 5 themes — Auto (follows system), Dark, Light, Dracula, Synthwave, Matrix
• 9 languages — English, Svenska, Deutsch, Español, Français, 日本語, Português (Brasil), Русский, 中文 (简体)
• Adjustable font size — 10–20px with live preview

Input & Media

• Voice input — Dictate messages using Web Speech API
• Drag & drop — Drop files directly onto the chat to attach them
• File upload — Attach images, PDFs, code files (.py, .txt)
• Screenshot capture — Take and attach screenshots of the current page
• Copy page text — Grab all text from the current tab as context

Developer Features

• Syntax highlighting — Code blocks are highlighted via highlight.js
• Ollama model management — List, pull, and delete local Ollama models from the settings panel
• Token counter — Estimated token count shown per conversation
• Custom endpoints — Set your own Ollama and Stable Diffusion URLs
• Keyboard shortcuts — Ctrl+Enter to send, Ctrl+N for new chat, Escape to close modals

Configuration

1. Click the ⚙️ Settings button
2. Add your API keys for the providers you want to use
3. Get free or paid API keys from each provider's website (links included in settings)

Free/Cheap Options

• Google Gemini — Free tier available at aistudio.google.com
• Groq — Free tier with generous limits at console.groq.com
• DeepSeek — Very affordable at platform.deepseek.com
• Ollama — Run models locally for free (requires Ollama installed)
• OpenRouter — Pay-per-use access to many models

Supported Providers

| Provider | Chat | Images | Notes |
|----------|------|--------|-------|
| OpenAI | ✅ | ✅ | GPT-4o, o1, o3, etc. |
| Anthropic | ✅ | — | Claude 3/4, Opus, Sonnet, Haiku |
| Google Gemini | ✅ | ✅ | Gemini 2.5 Flash, 3.1 Flash, Imagen 4 |
| DeepSeek | ✅ | — | V3, R1 reasoning models |
| Grok (xAI) | ✅ | ✅ | Grok 2, Imagine |
| Groq | ✅ | — | Ultra-fast inference |
| Mistral | ✅ | — | Mistral, Mixtral, Codestral |
| Cohere | ✅ | — | Command R/R+, Aya |
| Perplexity | ✅ | — | Online search-enabled models |
| Hugging Face | ✅ | — | 200+ models via Inference API |
| Ollama | ✅ | — | Local models (Llama, Mistral, etc.) |
| Ollama Cloud | ✅ | — | Hosted Ollama |
| OpenRouter | ✅ | — | Aggregator for 200+ models |
| Together AI | ✅ | ✅ | FLUX, Llama, Qwen, DeepSeek |
| Fireworks AI | ✅ | ✅ | FLUX, Llama, Qwen |
| DeepInfra | ✅ | ✅ | FLUX, Bria, Seedream |
| Novita AI | ✅ | ✅ | FLUX, Qwen-Image, Seedream |
| SiliconFlow | ✅ | ✅ | FLUX, DeepSeek, Qwen |
| Hyperbolic | ✅ | ✅ | FLUX |
| Alibaba | ✅ | — | Qwen models |
| SambaNova | ✅ | — | Llama, DeepSeek |
| NVIDIA | ✅ | — | Nemotron |
| ChatAnywhere | ✅ | — | GPT-4o proxy |
| Kilo AI | ✅ | — | Claude, GPT via proxy |
| Zhipu AI | ✅ | — | GLM models |
| Vercel | ✅ | — | AI Gateway |
| zAI | ✅ | — | Various models |
| Zo.computer | ✅ | — | Zo Computer assistant |
| Stable Diffusion | — | ✅ | Local A1111/Forge API |

Tech Stack

• Manifest V3 browser extension
• Vanilla JavaScript (ES modules)
• CSS custom properties (theming)
• highlight.js for syntax highlighting
• Web Speech API for voice input