Qwen3-based text embedding model with strong multilingual retrieval performance. Output dimension 4096; supports the `dimensions` parameter to shorten output.
LLM Gateway routes requests to the best providers that are able to handle your prompt size and parameters.