G4F Provider & Model Tester

A minimal utility to benchmark free g4f providers across several models and save results to g4f_test_results.json. Use it to discover which provider+model pairs are currently working and the fastest.

What it does

• Iterates over available g4f.Provider entries and tests a small set of models.
• Runs tests concurrently and prints progress with successes/errors.
• Saves a full report to g4f_test_results.json with timings and a preview.

Requirements

• Python 3.10+

Installation

Option A — global/default Python

pip install -U g4f

If pip belongs to a different interpreter, use:

python -m pip install -U g4f

On Windows with the launcher:

py -m pip install -U g4f

Option B — isolated virtual environment

Windows (PowerShell):

py -m venv .venv
.venv\Scripts\Activate.ps1
python -m pip install -U pip
python -m pip install -r requirements.txt

Linux/macOS:

python3 -m venv .venv
source .venv/bin/activate
python -m pip install -U pip
python -m pip install -r requirements.txt

Run

Global/default Python:

py test.py

Virtual environment (Windows):

.venv\Scripts\python test.py

Virtual environment (Linux/macOS):

.venv/bin/python test.py

Output

• The script writes a detailed report to g4f_test_results.json:
  • all_results: every tested pair with status, time, and preview.
  • successful: convenience list of successful pairs.
  • success_rate and totals.

Troubleshooting

• ModuleNotFoundError: No module named 'g4f'
  • Install the dependency in the interpreter you are using (see Installation).
• Lots of provider errors/logs
  • Normal: many providers are region/ratelimit/credential dependent. Focus on the SUCCESS items.
• Unclosed client session warnings
  • Cosmetic, from upstream providers. You can ignore or reduce concurrency in test.py.
• Proxy

  • If you need a proxy, export HTTPS_PROXY or HTTP_PROXY before running.

  • g4f, ai, providers, openai, chatgpt, python

What is this script?

test_providers_g4f.py is a benchmark/helper that probes g4f providers to find currently working combinations. You can use the generated JSON to hardcode a priority list in your own plugins/bots.