jina-reranker-v3.5 · Jina AI
jina-reranker-v3.5 is a 0.6B-parameter multilingual listwise document reranker and a drop-in upgrade to jina-reranker-v3. It keeps the same last-but-not-late (LBNL) listwise interface — ranking a query against many candidates in one forward pass — while improving domain robustness, structured-data ranking, multilingual retrieval, and inference efficiency.
jina-reranker-v3.5 is a 0.6B-parameter multilingual listwise document reranker and a drop-in upgrade to jina-reranker-v3. It keeps the same last-but-not-late (LBNL) listwise interface — ranking a query against many candidates in one forward pass — while improving domain robustness, structured-data ranking, multilingual retrieval, and inference efficiency.
Jina Reranker V3.5 has a 131,000 token context window.
On AIHubMix, Jina Reranker V3.5 costs $0.05 per million input tokens and $0.05 per million output tokens.
Jina Reranker V3.5 accepts text and image input.
Jina Reranker V3.5 is available through the AIHubMix unified API. The API is OpenAI-compatible: point your OpenAI SDK at https://aihubmix.com/v1, use your AIHubMix API key, and set the model name to jina-reranker-v3.5 — no other code changes needed.
Jina Reranker V3.5 is developed by Jina AI. AIHubMix aggregates it alongside models from other providers behind one API and one bill.
A 3.8-billion-parameter general vector model (embedding model) for state-of-the-art…
A 3.8-billion-parameter general vector (embedding) model providing state-of-the-art…
DeepSearch combines search, reading, and reasoning capabilities to pursue the best…
A general-purpose vector model with 3.8 billion parameters, used for multimodal and…
Multimodal multilingual document reranker, 131K context, 0.6B parameters, for visual…
Multi-modal Embeddings Model, multilingual, 1024-dimensional, 865M parameters.
Use Jina Reranker V3.5 via the AIHubMix unified API — one interface for every major LLM.