RVNLLM
Open Source • High Performance

Lightning-fast LLM toolingbuilt for developers

Transparent, and secure. Analyse and introspect any model in any format with confidence using our high-performance toolkit designed by developers, for developers. (Open source version only includes the GGUF format)

Built for Performance

Every component optimized for speed, security, and developer experience.

Lightning Fast
Optimized Rust core with zero-copy operations and intelligent caching
  • • Sub-100ms response times
  • • Memory-efficient processing
  • • Parallel request handling
Multi-Model Support
Unified API for all major LLM providers with intelligent routing (pro).
  • • OpenAI, Anthropic, Cohere
  • • Local model support
Universal File Support
Native support for all major model formats and file types
  • • GGUF (OSS, pro), SafeTensors (pro), PyTorch (pro)
  • • Hugging Face models
  • • Auto-format detection

Simple. Powerful. Fast.

Get started with just a few lines of code

import rvnllm

model = '70b-model.gguf' 
    
df = rvn_py.info(model) 
    .filter(pl.col("size_bytes") > 1_000_000) 
    .select(["tensor", "shape", "dtype"]) 
    .sort("size_bytes") 
    .collect()