Bhumi Logo

Bhumi भूमि

The blazing-fast AI inference client for Python

Built with Rust for performance. Unified interface for OpenAI, Anthropic, Gemini, Groq, and more.

$pip install bhumi
View on GitHubDocumentation
3x
Faster
60%
Less Memory
4+
AI Providers

Try Different Providers

Same simple API, multiple AI providers

OpenAI GPT-4

Most popular AI model

1from bhumi.base_client import BaseLLMClient, LLMConfig
2import asyncio
3
4async def main():
5    config = LLMConfig(
6        api_key="YOUR_OPENAI_API_KEY", 
7        model="openai/gpt-4o",
8        max_tokens=1000
9    )
10    
11    client = BaseLLMClient(config)
12    
13    response = await client.completion([
14        {"role": "user", "content": "Explain quantum computing"}
15    ])
16    
17    print(response['text'])
18
19asyncio.run(main())

Quick Start

Get started with your first AI call in seconds

Example: OpenAI GPT-4

Simple async client usage

1from bhumi.base_client import BaseLLMClient, LLMConfig
2import asyncio
3
4async def main():
5    config = LLMConfig(
6        api_key="YOUR_OPENAI_API_KEY", 
7        model="openai/gpt-4o",
8        max_tokens=1000
9    )
10    
11    client = BaseLLMClient(config)
12    
13    response = await client.completion([
14        {"role": "user", "content": "Hello, world!"}
15    ])
16    
17    print(response['text'])
18
19asyncio.run(main())

Performance

Benchmarked against popular alternatives

Performance Comparison
Response time comparison across different AI clients(Bhumi is 3.2x faster than alternatives)

Why Bhumi?

Built for developers who need reliable, fast AI inference

Lightning Fast
2-3x faster than alternatives with optimized Rust implementation
Multi-Model Support
Seamless integration with OpenAI, Anthropic, and Gemini models
Resource Efficient
Uses 60% less memory while handling concurrent requests
Production Ready
Battle-tested in high-throughput environments with 99.9% uptime
Open Source
Apache 2.0 licensed, free for commercial use with attribution
Enterprise Ready
Built-in rate limiting, error handling, and monitoring

Fast

Rust-powered performance with async Python interface

Simple

Unified API for all major AI providers

Reliable

Built-in rate limiting and error handling

Ready to go faster?

Install Bhumi and make your first AI call in under 2 minutes