Not a leaderboard. Models, pricing, licenses, availability and features change. These official starting points are not all live-verified.
| Organization | Focus | Official starting point |
|---|---|---|
| OpenAI | General/reasoning/multimodal models and APIs | https://developers.openai.com/ |
| Anthropic | Claude, prompting/context and APIs | https://docs.anthropic.com/ |
| Gemini/multimodal | https://ai.google.dev/gemini-api/docs | |
| Meta | Llama open-weight models | https://www.llama.com/ |
| Mistral | Commercial/open-weight | https://docs.mistral.ai/ |
| DeepSeek | Language/reasoning/coding | https://api-docs.deepseek.com/ |
| Alibaba Qwen | Multilingual/coding/multimodal/open-weight | https://qwenlm.github.io/ |
| xAI | Grok | https://docs.x.ai/ |
| Cohere | Enterprise/retrieval | https://docs.cohere.com/ |
| Amazon Bedrock | Managed multi-model access | https://docs.aws.amazon.com/bedrock/ |
| Microsoft Azure AI | Model catalog/enterprise | https://ai.azure.com/ |
| NVIDIA | Models/infrastructure | https://build.nvidia.com/ |
| Moonshot AI | Kimi | https://platform.moonshot.ai/ |
| Z.ai | GLM | https://docs.z.ai/ |
| MiniMax | General/multimodal | https://www.minimax.io/ |
How to compare
Compare candidates on the same task: capability, latency, cost, context, modalities, tools, privacy, retention, deployment, licensing, accessibility, and evaluations. No provider is universally best. Verify official sources and record date/model version.