One Interface
for
Every AI Model

Switch between Google, OpenAI, Anthropic, or any AI provider without changing your code. Open source, simple, free.

Python code showing how to use ProxAI with different models

Integrated with all major AI providers

DeepSeek
And more...

ProxAI in Action

Write once, connect to any model. Start building in minutes.

Screenshots of the dashboard project showing desktop version

One Line Setup

Just import the library and you're ready to go. No complex configuration required.

import proxai as px

Switch Models Instantly

Change between any provider or model with a single line of code.

px.set_model(('openai', 'gpt-4'))
px.set_model(('claude', 'sonnet'))
px.set_model(('gemini', 'flash'))

Multi-Model Experiments

Run the same prompt across multiple models to compare outputs.

for model in px.list_models():
px.set_model(model)
result = px.generate_text(...)

Complete Solution

ProxAI works with all the features you expect from modern AI APIs:

Text Generation

System Messages

Temperature

Chat Completions

Token Control

Function Calling

Philosophy

Model-agnostic AI development

The AI landscape is constantly evolving. Don't tie your application to a single provider. Stay flexible with ProxAI.

Zero Vendor Lock-in

Switch between Anthropic, OpenAI, Google, or any provider with a single line of code change.

px.set_model(('anthropic', 'claude-3'))

Instant Experimentation

Compare outputs between different models and providers in minutes, not days.

for model in px.list_models("large"):

Future-proof

New models get added automatically. Your code stays clean regardless of API changes.

Auto-update model registry

Full Control

Use our API keys for testing or add your own. Run completely on-prem or in your cloud.

Self-hosted option available

ProxDash: AI Analytics

Track all your AI experiments in one place. Monitor usage, costs, and performance across providers.

Cost Optimization

Track token usage and costs across providers. Identify the most cost-efficient models for your use case.

Performance Comparison

Compare response quality, latency, and reliability across models to find the best fit for your application.

Experiment Management

Organize and track experiments with custom tags. Create shareable reports for your team.

Advanced Logging and Tracing

Comprehensive request and response logging with full metadata.

ProxDash analytics dashboard showing cost features

Compare Models Side-by-Side

Stop guessing which AI model is best for your use case. Test them all with the same code.

# Define your function once
def summarize(article_text):
return px.generate_text(
prompt=f"Summarize this: {article_text}",
max_tokens=100)
# Sample article to summarize
article = "The latest AI developments have..."
# Compare across models
results = {}
for model in px.list_models():
px.set_model(model)
results[model] = summarize(article)
results[model + "_time"] = px.last_call_time
# Results
('claude', 'sonnet'): 0.8s
('openai', 'gpt-4'): 1.2s
('gemini', 'gemini-1.5-pro'): 0.9s

Objective Benchmarking

Compare response quality, speed, and token usage across all providers to make data-driven decisions.

A/B Testing

Run production A/B tests with different models to find the optimal balance of cost and performance.

Cost Optimization

Find the most cost-effective model for each task. Stop overpaying for capabilities you don't need.

100% Open Source

ProxAI is and will always remain completely open source. Join our community of builders and contributors.

GitHub Repository

Star our repo, submit issues, or contribute code. Every contribution helps.

Join Our Community

Connect with other developers on our Discord, share your use cases, and get help with your projects.

Roadmap & Updates

See what's coming next and help shape the future of ProxAI with feature requests.

Why Open Source Matters

Transparency

See exactly how your AI requests are handled and ensure no data is misused.

Innovation

Community contributions bring new features and integrations faster.

Reliability

No lock-in. The code is yours to use.

Extensibility

Build on top of ProxAI, customize it to your needs, and share improvements.

Ready to Start?

ProxAI is simple to use. Get your first AI response in 2 minutes.