Python · Instant Download

GenAI Multi-Provider
Starter Kit

Claude Gemini ChatGPT Grok

Stop rewriting AI boilerplate for every provider. One clean Python codebase runs all four — against real files, with four built-in operations. Plug in your API keys and run in 5 minutes.

AI Providers

File Types

Actions

5 min

To Running

$20

one-time · instant download
57.6 KB · Python 3.9+

Get Instant Access →

⚡ Instant Download 🔑 Own your API keys 🔧 Fully extensible

Up and Running

From zero to running AI on real files in 5 minutes.

Step 1

pip install -r requirements.txt

One command installs all dependencies

Step 2

Add your API keys to .env

Claude, Gemini, ChatGPT, or Grok — use whichever you have

Step 3

python run.py

Menu-driven: pick provider → file → action → run

Step 4

Select Provider
Select File
Select Action
Run

Interactive menu, no config files to edit

What's Inside

Clean architecture built for extension.

Project Structure

run.py← single entry point. pick provider, file, action.

actions/← 4 JSON prompt definitions

├── extract.json

├── summarize.json

├── parse.json

└── explain.json

providers/← 4 Python adapters (~15 lines each)

├── claude.pyclaude-sonnet-4-5

├── gemini.pyauto-detects model

├── chatgpt.pygpt-4o

└── grok.pygrok-3-latest

sample-files/← real files to test against immediately

├── us-constitution.pdf

├── ev-population.csv

└── powerball.json

requirements.txt

🔌

4 AI Providers, One Interface

Claude, Gemini, ChatGPT, and Grok all run through the same menu. Swap providers in a single keypress — no code changes required.

📄

Real Files Included

3 sample files ship with the kit — US Constitution PDF, EV Population CSV, and a Powerball JSON. Start testing intelligence on real data immediately.

⚙️

4 Built-In Actions

Extract key facts, summarize content, parse structured data, and explain concepts — each defined as a clean JSON prompt in the actions/ folder.

🧩

Drop-In Extension

Add a new provider: one Python file in providers/. Add a new action: one JSON file in actions/. Zero refactoring. The architecture scales without changing the core.

🗂️

PDF · CSV · JSON Support

Three of the most common document formats — all handled out of the box. Swap in your own files by changing a single path in run.py.

⚡

5-Minute Setup

pip install, add your API keys, run python run.py. The menu does the rest. Built for developers who want to start comparing providers, not configuring infrastructure.

Confirmed Working Models

All four providers. All confirmed working.

Claude

claude-sonnet-4-5

Anthropic's production Sonnet model. Strong at document understanding and structured extraction.

Gemini

auto-detects

Automatically detects the best available model for your account. No manual version pinning needed.

ChatGPT

gpt-4o

OpenAI's flagship multimodal model. Consistent, well-documented, and widely benchmarked.

Grok

grok-3-latest

xAI's latest model. Useful for comparing reasoning and style against the other three providers.

4 Actions

🔍 Extract 📋 Summarize </> Parse 💡 Explain

3 File Types

📕 PDF 📊 CSV 📋 JSON

Built for Developers

Extend it in minutes, not hours.

➕

Add a New Provider

Drop one Python file in the providers/ folder. Follow the same ~15-line adapter pattern the existing four use. That's it — the menu picks it up automatically.

# providers/my_provider.py def run(prompt, content): client = MyProviderClient(api_key) return client.complete(prompt + content)

🎯

Add a New Action

Create one JSON file in the actions/ folder with your prompt definition. Name it, describe it, and it appears in the menu on the next run. No refactoring needed.

{ "name": "Translate", "prompt": "Translate the following content to Spanish, preserving all structure and meaning:" }

📁

Use Your Own Files

The three sample files are starting points. Swap in any PDF, CSV, or JSON by changing the file path in run.py — or extend the file picker to browse a directory.

# In run.py — change one line FILE_PATH = "path/to/your/file.pdf"

🔬

Compare Providers Side-by-Side

Run the same action on the same file across all four providers in sequence. The identical prompt structure makes output comparison clean and reproducible.

# Run all providers on the same file for provider in ["claude", "gemini", "chatgpt", "grok"]: result = run_provider(provider, file, action)

Why This Kit

vs. building it yourself

Capability	Build From Scratch	This Kit ✓
4 providers wired up	✗ hours of docs	✓ included
PDF / CSV / JSON parsing	✗ research & install	✓ included
Prompt abstraction layer	✗ design from scratch	✓ JSON-based
Menu-driven runner	✗ build your own CLI	✓ run.py
Sample files to test with	✗ source yourself	✓ 3 included
Extension pattern documented	✗ figure it out	✓ drop-in
Time to first working demo	✗ hours to days	✓ 5 minutes

GenAI Multi-Provider Starter Kit