Alternatives to Grouped Query Attention (GQA)

Japan-focused efficient AI models using bio-inspired evolution to build capable, compact foundation models.

Designed for institutional investors, Reflexivity combines trusted data from S&P Global, LSEG Datastream, Cboe, Nasdaq, and more with explainable AI to surface insights - not noise.

Foundation models for simulating Earth's systems, starting with weather.

Connects LLMs and AI agents to real-time mainframe data for enterprise insights.

Developing foundational AI models and the infrastructure to train and deploy them.

Arcee provides U.S.-made, open source large language models and checkpoints for insight into raw model intelligence.

Mistral AI provides frontier AI models, assistants, and agents with compute infrastructure and services.

Chinese AI startup developing advanced large language models and AI agents, including the Kimi chatbot, with a focus on AGI research.

AI-powered video editing platform that transforms long-form content into short, engaging social media clips.

High-performance AI inference platform built by former Meta PyTorch engineers.

AI-powered research assistant for scientific literature review and data extraction.

Etched is building the hardware for superintelligence by burning the transformer architecture into their chips.

Developing a novel generative AI model with a subquadratic architecture for significantly increased context windows.

A unified AI model aggregation and distribution gateway that converts various large language models into OpenAI, Claude, and Gemini compatible interfaces.

AI development platform for building and deploying AI applications.

AI computing innovator revolutionizing edge computing solutions with programmable processors and software platforms.

Kuaishou's generative AI service for creating videos from text and image descriptions.

Fermi is a startup developing a novel approach to AI model training and deployment.

AI-powered legal research and analysis platform for legal professionals.