Prometheus AI Orchestrator

Send one question to multiple AI chatbots simultaneously. Collect all answers. Let an expert AI synthesize the best response.

A Chrome extension (Manifest V3) that orchestrates multiple AI services in parallel — ChatGPT, DeepSeek, Gemini, Grok, Claude, Qwen, and more. No API keys required for browser-tab mode.

Why

Different AI models excel at different tasks. Instead of manually copy-pasting your question across 5 tabs, Prometheus does it in one click and aggregates the results through a designated "expert" model.

Use cases:

• Compare answers from multiple LLMs side by side
• Get a synthesized "best of all" answer via expert aggregation
• Fact-check one AI with several others
• Research complex topics with diverse AI perspectives

Supported AI Services

Browser Tab Mode (no API keys needed)

Service	Status	URL
ChatGPT	Tested	chat.openai.com
DeepSeek	Tested	chat.deepseek.com
Gemini	Tested	gemini.google.com
Grok	Tested	grok.com
Qwen	Tested	chat.qwen.ai
Claude	Supported	claude.ai
Perplexity	Supported	perplexity.ai
Mistral Chat	Supported	chat.mistral.ai
Copilot	Supported	copilot.microsoft.com
HuggingChat	Supported	huggingface.co/chat
YandexGPT	Supported	ya.ru/ai
GigaChat	Supported	gigachat.ai

API Mode (free tiers)

Provider	Model	Free Tier
Gemini API	gemini-2.0-flash	Yes
Groq	llama-3.3-70b-versatile	Yes (1000 req/min)
OpenRouter	llama-3.3-70b, qwen3-235b	Yes
Mistral API	mistral-small-latest	Yes

How It Works

You type a question
        |
        v
  +-----------+
  | Prometheus |---> ChatGPT tab  ---> Answer 1
  |  (Chrome   |---> DeepSeek tab ---> Answer 2
  | Extension) |---> Gemini tab   ---> Answer 3
  |            |---> Grok tab     ---> Answer 4
  +-----------+
        |
        v
  All answers collected
        |
        v
  Expert AI synthesizes
  the final response

1. Scan — detects all open AI service tabs in your browser
2. Pre-warm — activates each tab sequentially to prevent throttling
3. Send — injects your question into each service's input field
4. Poll — monitors response status every 10 seconds
5. Extract — reads the generated answer from each service's DOM
6. Aggregate — sends all collected answers to a designated expert for synthesis

Installation

1. Clone or download this repository
2. Open chrome://extensions/
3. Enable Developer mode (top-right toggle)
4. Click Load unpacked
5. Select the project folder (containing manifest.json)

Usage

1. Open tabs with your preferred AI services (ChatGPT, DeepSeek, Gemini, etc.)
2. Click the Prometheus icon in Chrome toolbar
3. Type your question
4. Select which AI assistants to query and which one is the expert
5. Click Send
6. Wait for all responses to arrive
7. Review individual answers or the expert's aggregated synthesis

Project Structure

prometheus-ai-orchestrator/
├── manifest.json           # Chrome Manifest V3
├── background/
│   └── service-worker.js   # Orchestration, polling, tab management
├── content/
│   └── content.js          # AI service adapters (DOM injection & extraction)
├── popup/
│   ├── popup.html          # Extension popup UI
│   ├── popup.js            # UI logic & state management
│   ├── popup.css           # Styles
│   └── markdown.js         # Markdown rendering for responses
└── icons/                  # Extension icons (16/32/48/128px)

Key Features

• Zero-config browser mode — just open AI tabs and go, no API keys needed
• 12 AI services supported (ChatGPT, Claude, Gemini, Grok, DeepSeek, Qwen, Perplexity, Mistral, Copilot, HuggingChat, YandexGPT, GigaChat)
• 4 API providers with free tiers (Gemini, Groq, OpenRouter, Mistral)
• Expert aggregation — one AI synthesizes all others' answers
• Parallel execution — all services queried simultaneously
• Manifest V3 — modern Chrome extension architecture
• No backend server — everything runs locally in your browser

Roadmap

• Firefox support (WebExtensions API)
• Configurable polling intervals
• Response quality scoring
• Export results (JSON/Markdown)
• Custom expert prompts
• Side-by-side diff view

Contributing

PRs welcome. Please open an issue first to discuss significant changes.

License

Apache-2.0