Pruna AI

Pruna AI is an AI optimization engine that provides a frictionless solution to optimize and compress machine learning models, making...

On April 29, 2025, Pruna AI announced a collaboration with Koyeb to achieve 5x faster inference speeds on serverless GPUs...

Pruna AI now supports Hugging Face bitsandbytes with PyTorch torch.compile and has fixed several bugs...

Pruna AI now offers the ability to speed up and quantize any diffusion model and evaluate with CMMD using EvaluationAgent, while running Flux models with half the memory...

Pruna AI released ComfyUI-Pruna Nodes, enhancing image processing speed by 2x on SD models and 25% on Flux...

Pruna AI has the largest collection of over 9,000 models on Hugging Face, including models from Mistral AI, DeepSeek AI, and Meta...

Pruna AI is participating in the AI summit week with events from February 6-12, including the AI Action Summit and Sustainable AI Networking Happy Hour...

Caroline Chavier, a Business Angel, supports Pruna AI, which optimizes AI models to be faster, cheaper, smaller, and greener, reducing carbon emissions by up to 91%...

Pruna AI successfully evaluated five base models for a major German car manufacturer, achieving significant speed improvements across various hardware setups, with Gaspar Rochette and Louis Leconte leading the project...

Pruna AI collaborated with Hugging Face to create SMOLLM2, a model that is now 7 times smaller and 2 times faster...