Model Tools & Dev Platform

Frontier models deployed through everyday products

OpenAI researches and deploys advanced AI through ChatGPT, developer APIs, and Codex, connecting foundation models with consumer, enterprise, and software development workflows.

AI that manages the infrastructure bill

North automates cloud, AI, and data spending, combining cost controls and optimization with a conversational FinOps agent that answers questions and builds dashboards.

Frontier models and a deployable enterprise AI stack

Mistral develops frontier AI models, developer tools, applications, and compute infrastructure, helping enterprises build customized systems across language, coding, document, and speech workflows.

From open-source frameworks to owned GPU compute

Lightning AI combines PyTorch Lightning developer tools with owned GPU infrastructure, providing one platform to build, train, deploy, and scale AI systems.

Understanding how a voice interaction feels

Hume builds audio-native AI models that turn emotion, tone, timing, and conversational context into structured judgments for evaluating and improving voice interactions.

The open collaboration hub for machine learning

Hugging Face helps developers discover, share, evaluate, and deploy AI models through an open community spanning models, datasets, applications, and libraries.

Enterprise AI built around business decisions

Fractal develops enterprise AI platforms and decision systems, combining engineering, analytics, and domain expertise to deploy models and agents into business operations.

Open models with deployment control

Reflection develops open-weight models and supporting software for coding, reasoning, and agents, giving organizations deployment control through its API or their own infrastructure.

Data and evaluations for reliable AI

Scale AI provides training data, model evaluations, and enterprise AI applications, helping organizations build and deploy systems whose capabilities can be measured.

Track, evaluate, and improve the AI development lifecycle

Weights & Biases gives AI teams tools to track experiments, evaluate models and agents, inspect application traces, and improve deployed systems through shared evidence.

Specialized models served as production AI infrastructure

Fireworks helps teams train and serve specialized AI models, adapting open models to customer data and workflows across text, image, audio, and multimodal applications.

Neural optimization for database compute

Espresso AI uses machine learning to optimize database queries, compute allocation, and scheduling, reducing warehouse costs without requiring engineers to tune infrastructure manually.

Specialist language AI for global communication

DeepL develops language AI for translation, writing, and voice, helping businesses communicate across languages through specialist models and enterprise deployment controls.

A shared platform for enterprise AI

Dataiku brings data preparation, machine learning, generative AI, and governance into one collaborative platform so business and technical teams can build AI together.

Enterprise AI becomes more useful when models, agents, transactional databases, and analytics share governed access to the same business data

Databricks unifies enterprise data, analytics, governance, and AI development, helping teams build agents and applications grounded in their own business data.

Codebase-aware agents for building software

Cursor builds an AI coding environment where developers delegate work to agents grounded in their codebase, combining model research with editing and software delivery tools.

Purpose-built cloud infrastructure for AI workloads

CoreWeave provides GPU cloud infrastructure and software for AI training, inference, and model development, combining compute, storage, networking, and operational tooling at scale.

Organizations can adopt advanced AI while retaining control over deployment, data access, and governance in their own cloud or on-premises infrastructure

Cohere builds enterprise AI models and platforms for search, retrieval, and agents, with private deployments that keep institutional data within customer-controlled environments.

Software agents that carry engineering tasks through delivery

Cognition builds software agents that plan, write, test, and ship code, with Devin working inside engineering teams’ repositories and existing development tools.

Companies can run specialized AI models in production without assembling inference, optimization, deployment, and observability infrastructure themselves

Baseten provides infrastructure and embedded engineering for deploying, optimizing, and serving AI models, with training and inference across cloud, hybrid, and self-hosted environments.

Observe, evaluate and improve production AI

Arize provides AI observability and evaluations that trace, test, and improve models and agents from development through production, including its open-source Phoenix tools.

Claude and the science of reliable AI

Anthropic develops Claude models, assistants, and coding tools, combining frontier AI research with an explicit focus on reliable, interpretable, and steerable systems.