Up to date
Models › jina-reranker-v2 family
jinaai CC BY-NC 4.0 out 2 years ago
278M parameters, a query and passages in, relevance scores out, 1K tokens of context.
Verified 2026-10-25 Rank #53 in Embeddings · score 88.8 method v1 Sources 3 Downloads 30d 1.5M · #83 of 619 tracked
Picked by rule from 6 tracked pieces: the lab's own post, the most shared, a field report, the latest.
jina-reranker-v2-base-multilingual is a 278M-parameter cross-encoder reranker supporting 100+ languages with a 1,024-token context window (524,288 tokens with Flash Attention 2).
The 1 pick above, plus these 5. Top is the radar's score: discussion, source authority and the lab's own posts, fading with age.
Hugging Face does not say who downloads a model or with what. These are the proxies: API providers listing it, quantized builds, fine-tunes/adapters/merges, and Spaces that load it.
No API provider serves it, so its downloads are people running it themselves: 10 quantized builds and 41 Spaces.
Top Spaces running it: Ekimetrics/climate-question-answering ♥ 150 · TheresaQWQ/jina-rerank ♥ 1 · UniLawChat/UniLawChat ♥ 1 · Hamdy005/Study-Buddy ♥ 1
Weights only, an estimate: leave room for the KV cache. Method
About 0.2 GB at INT4, 0.3 GB at FP8, 0.7 GB at BF16: weights only, so leave room for the KV cache.
Estimate params × bytes per param × 1.2 · how we estimate
Weights, every dated article, and repo commits on one rail.
Teal is the weights (2024-06-19), amber triangles are 1 dated article (each links to it), grey ticks are repo commits.
38 derivatives on the Hub. Which build to run, by runtime: the most downloaded in each class.
From the Hugging Face model tree; quantized builds in teal, fine-tunes, merges and adapters in greys.
Everything else, folded: the full tables, every source.
jina-reranker-v2-base-multilingual is a 278M parameter transformer-based cross-encoder designed for multilingual text reranking tasks.[1][2] The model supports a context length of up to 1024 tokens and is released under a CC BY-NC 4.0 license.[3][4]
Generated 2026-10-07 15:04 UTC by parasail-gemma-4-26b-a4b-it; each numeral points at the fact it rests on. Not a verdict.
Labels are evidence, not verdicts: measured a third-party leaderboard row for this model; author-reported a number the model author published in the card or eval files; structural a deterministic signal from the name, config or chat template, or a decision-model estimate. Each chip links to its source. All use cases.
Listed in Embeddings
| Pipeline tag | text-ranking |
|---|---|
| Library | transformers |
| Architectures | XLMRobertaForSequenceClassification |
| Dtype | bfloat16 |
| Params by dtype | BF16: 278,437,633 |
| Size band | S |
| MoE | no |
| Context length | 1,026 (config.json max_position_embeddings) |
| Languages | multilingual |
| License | CC BY-NC 4.0 (cc-by-nc-4.0) |
| Hub snapshot | 2026-10-25 |
vLLM: supported · XLMRobertaForSequenceClassification · 7050ef881ff0835f6890a47db1859a24c65481d6SGLang: supported · XLMRobertaForSequenceClassification · 773cb938fe9484a8f7e7e46fb1b1fe96412c59ed
vLLM registry (7050ef881ff0835f6890a47db1859a24c65481d6) · SGLang registry (773cb938fe9484a8f7e7e46fb1b1fe96412c59ed) · Matched by architecture class name against each project's registry on main; 'supported' means the architecture is registered, not that this checkpoint was tested.
Weights-only memory for jina-reranker-v2-base-multilingual: params × bytes/param × 1.2 overhead. KV cache and activations are extra. How this is computed.
| BF16 (2 bytes/param) | 0.7 GB · Edge (<4 GB) |
|---|---|
| FP8 (1 byte/param) | 0.3 GB · Edge (<4 GB) |
| INT4 (0.5 bytes/param) | 0.2 GB · Edge (<4 GB) |
Weights only (params × bytes/param × 1.2 overhead); KV cache and activations extra.
finetune: 28quantized: 10
| Date | Commit | SHA |
|---|---|---|
| 2025-10-21 | fix: empty passage | 9cfeff2 |
| 2025-09-10 | Update model metadata to set pipeline tag to the new `text-ranking` and tags to `sentence-transformers` (#35) | 2f894e6 |
| 2025-01-06 | Update README.md | eed787b |
| 2024-09-26 | Adding `safetensors` variant of this model (#20) | 1267477 |
| 2024-07-17 | Update tokenizer.json for max length (#18) | 6371035 |
| 2024-06-28 | Describe CrossEncoder integration with Sentence Transformers (#8) | 2f34eeb |
| 2024-06-26 | Upload ONNX weights (+ Transformers.js integration) (#9) | 298e48c |
| 2024-06-26 | Update block.py (#10) | 8469b0a |
| 2024-06-25 | licence info and inference API (#7) | 20be9d6 |
| 2024-06-25 | fix: sagemaker import issue (#6) | 8bb94e2 |
| 2024-06-24 | docs: update benchmark (#5) | 6949f0a |
| 2024-06-24 | Update README.md | 75e0111 |
| 2024-06-24 | Update README.md | 1ae7f51 |
| 2024-06-24 | docs: update benchmark (#4) | 3aaa84b |
| 2024-06-21 | fix: markdown table | 667e9cb |
| 2024-06-21 | fix: markdown table | 7837e7d |
| 2024-06-21 | fix: markdown table | ce1a0bf |
| 2024-06-21 | docs: refine README for model releasing (#2) | a20b077 |
| 2024-06-20 | chore: update readme | 7c006e5 |
| 2024-06-20 | chore: update readme | 86ead8f |
| 2024-06-20 | chore: update readme | 80d2ebe |
| 2024-06-19 | fix: add missing module | eda82c8 |
| 2024-06-19 | fix: add missing module | 0be5eb2 |
| 2024-06-19 | fix: add missing module | 391704c |
| 2024-06-19 | fix: add missing module | 4c9340f |
| 2024-06-19 | fix: add missing module | 1d1bb09 |
| 2024-06-19 | chore: update readme | 182734b |
| 2024-06-19 | chore: update readme | 03c97a6 |
| 2024-06-19 | Upload folder using huggingface_hub | 67b96c4 |
| 2024-06-19 | initial commit | e72879e |
Ranked by source signals (HN, Lobsters, authority domains, official posts) and recency. Summaries are generated from a stored copy of the article text; every sentence cites a span of that text, and nothing is shown that failed the citation check.
Hub snapshot: 2026-10-25 · HF API: 2026-10-24 21:37 UTC · Providers: 2026-10-24 21:44 UTC · Articles: 2026-10-24 21:45 UTC · Summary: 2026-10-07 15:04 UTC · Last verified 2026-10-25
Email alerts are not open yet; the feeds above are the same data. See feeds for the item schema.
