DeepSeek-R1, DeepSeek-V3 and Phi-4 models, advanced LLM inference optimization
We now offer deepseek-ai/deepseek-R1 (671B in Public Preview status and distilled variants in GA), deepseek-ai/deepseek-V3 (Public Preview), and microsoft/Phi-4 models in Model Garden. Advanced LLM inference optimization techniques — prefix caching and speculative decoding — are available in select models in Model Garden in Public Preview.
Anthropic
Anthropic is an AI research and development company focused on building large-scale language models that are reliable, interpretable, and easier to control. They place a strong emphasis on AI safety and aligning their systems with human values.