Lightweight Qwen3 cross-encoder reranker with 32K context for latency-sensitive search and RAG applications. Fast scoring at minimal cost. 0.6B size.
LLM Gateway routes requests to the best providers that are able to handle your prompt size and parameters.