15 companies tracked in this category. Names and one-line descriptions are free to browse — pricing model, funding, founding year, headquarters, and cited sources are in the full dataset.
| Company | Description |
|---|---|
| AI21 Labs | Israeli AI company that develops the proprietary Jurassic large language models and the open-weight Jamba family, offered via AI21 Studio and major cloud/AI platforms. |
| Amazon Web Services (Bedrock) | Amazon Bedrock is AWS's fully managed generative AI service providing API access to foundation models from Amazon and third parties, including Amazon's own Titan and Nova model families. |
| Anthropic | AI safety-focused company that develops the Claude family of models, available through Claude apps and the Claude API. |
| Cohere | Canadian enterprise-focused AI company developing the Command family of large language models along with embedding and rerank models for secure/private enterprise deployment. |
| Databricks (Mosaic AI / DBRX) | Databricks is a data and AI platform company that acquired MosaicML in 2023 to form Mosaic AI, offering tools to train, fine-tune, and serve foundation models directly on customer data within the Lakehouse. |
| DeepSeek | Chinese AI lab, spun out of quantitative hedge fund High-Flyer, that develops open-weight foundation models including DeepSeek-V3 and the DeepSeek-R1 reasoning model. |
| Google DeepMind | Google's AI research and product lab that develops the Gemini family of multimodal models. |
| IBM (watsonx.ai / Granite) | IBM's foundation model business is built around the watsonx.ai enterprise AI development platform and the Granite family of open-weight, Apache 2.0-licensed language models. |
| Inflection AI | Develops large language models — including Inflection-2.5 — that power Pi, a consumer conversational AI assistant, and has since pivoted toward enterprise AI deployment. |
| Meta AI | Meta's AI division develops the open-weight Llama family of large language models plus multimodal and agentic AI research, distributed for self-hosting and integrated into Meta's apps. |
| Microsoft (Azure AI Foundry / Phi models) | Microsoft hosts third-party and OpenAI foundation models for enterprises through Azure AI Foundry, while Microsoft Research separately develops its own Phi family of small, efficient language models. |
| Mistral AI | French AI lab developing the Mistral/Mixtral family of open-weight and proprietary large language models, offered via its La Plateforme API and open downloads. |
| OpenAI | Develops and deploys frontier AI models, including the GPT-5 / GPT-4o family, accessible via ChatGPT and the OpenAI API. |
| Stability AI | UK-based AI company best known for its open text-to-image models Stable Diffusion, alongside StableLM language models and other generative media models. |
| xAI | Elon Musk-founded AI company developing the Grok family of large language models, integrated into the X platform; became a wholly-owned subsidiary of SpaceX following a 2026 all-stock merger. |