These are the embedding models currently available from NVIDIA.
Nvidia API KEY acquisition address: https://build.nvidia.com/settings/api-keys
Did you try? The NIM API is compatible with Discourse AI for LLM inference, so I’d guess the embeddings one may also be.