lm-arena-chatbot-arena

Autonomous mobility service enabling driverless operation on urban roads

AI data platform for annotating text, audio, images, video, and 3D data, plus model evaluation

An open-source hardware platform offering sensors, communication modules, and edge computing devices for prototyping.

A computer vision development platform for managing datasets, training models, and deploying inference services.

Browser extension that automatically captures screen activity and logs to generate defect reports

A developer productivity platform supporting structured code reviews and incremental change management

Airtrain AI is a user-friendly, no-code platform designed to streamline the large-scale fine-tuning, evaluation, and comparison of large language models (LLMs). The platform empowers AI developers, data scientists, and enterprises to customize open-source or commercial LLMs—including GPT-3.5, GPT-4, Claude, Gemini, and Llama 2—using their own datasets without writing any code. It supports large dataset exploration, semantic clustering, batch evaluation, and fine-tuning across multiple metrics, helping users meet domain-specific requirements while reducing AI costs by up to 90%. Airtrain AI also integrates with mainstream frameworks like LlamaIndex and provides a Playground for prompt testing and model response comparison.

Ollama is an open-source AI tool designed to run large language models locally on personal computers without relying on cloud services. It supports a variety of mainstream models, including Meta's Llama 3 and Mistral, enabling users to maintain full control over data privacy and model customization. By operating offline, Ollama reduces latency, enhances security, and lowers the costs associated with cloud-based AI usage. It is ideal for developers, researchers, and enterprises seeking AI solutions with high privacy and flexibility requirements, and can be seamlessly integrated into existing workflows or deployed in dedicated applications.

RunPod is a comprehensive AI cloud platform designed for machine learning and deep learning applications. The platform provides high-performance GPU and CPU resources to help users efficiently train, fine-tune, and deploy AI models. It supports containerized workloads and serverless computing, offering a balance of flexibility and cost-effectiveness.

Opal is a suite of tools designed by Google to help developers build safer and more responsible applications using LLMs. Integrated into Google AI Studio and Vertex AI, Opal provides configurable safety filters and evaluation capabilities to assess model outputs against safety policies and quality benchmarks before and during deployment.

Sakana AI is an innovative AI research and development company focused on creating transformative foundational models inspired by natural evolution and swarm intelligence. Its flagship product, AI Scientist, is the first fully automated system capable of independently conducting scientific research—from concept development and experimentation to paper writing and peer review. Sakana AI also advances AI high performance through tools like AI CUDA Engineer, achieving up to 100x faster AI training and inference speeds. Backed by investments from NVIDIA and leading Japanese enterprises, the company is committed to building a world-class AI laboratory and driving sustainable, efficient, and cutting-edge AI technologies to solve complex challenges.

MindSpore is a full-scenario deep learning framework developed by Huawei, designed to seamlessly integrate AI model development, training, and deployment across diverse environments including cloud, edge, and end devices. The framework supports a unified programming paradigm and is compatible with both dynamic and static computation graphs, enabling developers to balance debugging convenience with execution efficiency. MindSpore incorporates advanced optimization technologies such as automatic differentiation, model parallelism, and hardware acceleration (supporting GPU, NPU, Ascend, and others). Its modular architecture allows for multi-domain extensibility, and with built-in rich models and operators, it accelerates rapid prototyping and large-scale AI application development.