Pixtral 12B - Mistral vs ggml.ai

French AI startup Mistral has recently launched its first multimodal model, Pixtral 12B, which processes both images and text. Built on Mistral's Nemo 12B text model, Pixtral 12B boasts 12 billion parameters (24GB in size), enabling it to perform tasks like captioning images and counting objects within photos.

The model is available for download on GitHub and Hugging Face, under the Apache 2.0 license, allowing for unrestricted use, including fine-tuning and commercial applications.

Pixtral 12B is comparable to other multimodal models such as Anthropic’s Claude family and OpenAI’s GPT-4, capable of answering questions based on images using URLs or base64-encoded formats. While no web demos were available at launch, Mistral plans to roll out testing through its chatbot and API platforms soon.

This release follows the company's significant $645 million funding round, highlighting Mistral’s ambition to compete with major AI players by focusing on open models and corporate consulting services.

ggml.ai is at the forefront of AI technology, bringing powerful machine learning capabilities directly to the edge with its innovative tensor library. Built for large model support and high performance on common hardware platforms, ggml.ai enables developers to implement advanced AI algorithms without the need for specialized equipment. The platform, written in the efficient C programming language, offers 16-bit float and integer quantization support, along with automatic differentiation and various built-in optimization algorithms like ADAM and L-BFGS. It boasts optimized performance for Apple Silicon and leverages AVX/AVX2 intrinsics on x86 architectures. Web-based applications can also exploit its capabilities via WebAssembly and WASM SIMD support. With its zero runtime memory allocations and absence of third-party dependencies, ggml.ai presents a minimal and efficient solution for on-device inference.

Projects like whisper.cpp and llama.cpp demonstrate the high-performance inference capabilities of ggml.ai, with whisper.cpp providing speech-to-text solutions and llama.cpp focusing on efficient inference of Meta's LLaMA large language model. Moreover, the company welcomes contributions to its codebase and supports an open-core development model through the MIT license. As ggml.ai continues to expand, it seeks talented full-time developers with a shared vision for on-device inference to join their team.

Designed to push the envelope of AI at the edge, ggml.ai is a testament to the spirit of play and innovation in the AI community.