[GH-ISSUE #6922] Support for jinaai/jina-embeddings-v3 embedding model #4381

Open
opened 2026-04-12 15:19:09 -05:00 by GiteaMirror · 40 comments
Owner

Originally created by @sakthi-geek on GitHub (Sep 23, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/6922

jina-embeddings-v3 is a multilingual multi-task text embedding model designed for a variety of NLP applications. Based on the Jina-XLM-RoBERTa architecture, this model supports Rotary Position Embeddings to handle long input sequences up to 8192 tokens. Additionally, it features 5 LoRA adapters to generate task-specific embeddings efficiently.

Key Features:

  • Extended Sequence Length: Supports up to 8192 tokens with RoPE.
  • Task-Specific Embedding: Customize embeddings through the task argument with the following options:
    retrieval.query: Used for query embeddings in asymmetric retrieval tasks
    retrieval.passage: Used for passage embeddings in asymmetric retrieval tasks
    separation: Used for embeddings in clustering and re-ranking applications
    classification: Used for embeddings in classification tasks
    text-matching: Used for embeddings in tasks that quantify similarity between two texts, such as STS or symmetric retrieval tasks
  • Matryoshka Embeddings: Supports flexible embedding sizes (32, 64, 128, 256, 512, 768, 1024), allowing for truncating embeddings to fit your application.
Originally created by @sakthi-geek on GitHub (Sep 23, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/6922 jina-embeddings-v3 is a multilingual multi-task text embedding model designed for a variety of NLP applications. Based on the [Jina-XLM-RoBERTa architecture](https://huggingface.co/jinaai/xlm-roberta-flash-implementation), this model supports Rotary Position Embeddings to handle long input sequences up to 8192 tokens. Additionally, it features 5 LoRA adapters to generate task-specific embeddings efficiently. Key Features: - Extended Sequence Length: Supports up to 8192 tokens with RoPE. - Task-Specific Embedding: Customize embeddings through the task argument with the following options: retrieval.query: Used for query embeddings in asymmetric retrieval tasks retrieval.passage: Used for passage embeddings in asymmetric retrieval tasks separation: Used for embeddings in clustering and re-ranking applications classification: Used for embeddings in classification tasks text-matching: Used for embeddings in tasks that quantify similarity between two texts, such as STS or symmetric retrieval tasks - Matryoshka Embeddings: Supports flexible embedding sizes (32, 64, 128, 256, 512, 768, 1024), allowing for truncating embeddings to fit your application.
GiteaMirror added the model label 2026-04-12 15:19:09 -05:00
Author
Owner

@thiswillbeyourgithub commented on GitHub (Sep 24, 2024):

Also interested, here's their announcement post: https://jina.ai/news/jina-embeddings-v3-a-frontier-multilingual-embedding-model/

<!-- gh-comment-id:2371201706 --> @thiswillbeyourgithub commented on GitHub (Sep 24, 2024): Also interested, here's their announcement post: https://jina.ai/news/jina-embeddings-v3-a-frontier-multilingual-embedding-model/
Author
Owner

@krittaprot commented on GitHub (Sep 27, 2024):

Looking forward to this for English-Thai support as well!

<!-- gh-comment-id:2379096893 --> @krittaprot commented on GitHub (Sep 27, 2024): Looking forward to this for English-Thai support as well!
Author
Owner

@alew3 commented on GitHub (Oct 1, 2024):

+1

<!-- gh-comment-id:2385433456 --> @alew3 commented on GitHub (Oct 1, 2024): +1
Author
Owner

@antonio-lamberti commented on GitHub (Oct 2, 2024):

Is this available for Ollama? Still don't see it...

<!-- gh-comment-id:2387981575 --> @antonio-lamberti commented on GitHub (Oct 2, 2024): Is this available for Ollama? Still don't see it...
Author
Owner

@JoanFM commented on GitHub (Oct 2, 2024):

Hello,

This is not yet available. I will try to add the support in llama.cpp but not sure how much it may take.

<!-- gh-comment-id:2388513913 --> @JoanFM commented on GitHub (Oct 2, 2024): Hello, This is not yet available. I will try to add the support in llama.cpp but not sure how much it may take.
Author
Owner

@antonio-lamberti commented on GitHub (Oct 17, 2024):

Let's hope soon!!! Thanks for the great job anyway 💯

<!-- gh-comment-id:2419042973 --> @antonio-lamberti commented on GitHub (Oct 17, 2024): Let's hope soon!!! Thanks for the great job anyway 💯
Author
Owner

@basileaw commented on GitHub (Nov 4, 2024):

+1!

<!-- gh-comment-id:2455772738 --> @basileaw commented on GitHub (Nov 4, 2024): +1!
Author
Owner

@YikaiHu commented on GitHub (Nov 5, 2024):

+1

<!-- gh-comment-id:2456178809 --> @YikaiHu commented on GitHub (Nov 5, 2024): +1
Author
Owner

@alainlizotte commented on GitHub (Nov 5, 2024):

+1

Would like to be able to use this model !

<!-- gh-comment-id:2457653090 --> @alainlizotte commented on GitHub (Nov 5, 2024): +1 Would like to be able to use this model !
Author
Owner

@canwushuang commented on GitHub (Nov 11, 2024):

+1 臣附议

<!-- gh-comment-id:2467079014 --> @canwushuang commented on GitHub (Nov 11, 2024): +1 臣附议
Author
Owner

@beyondmehayri commented on GitHub (Nov 12, 2024):

+1

<!-- gh-comment-id:2471860763 --> @beyondmehayri commented on GitHub (Nov 12, 2024): +1
Author
Owner

@wahyusaputra commented on GitHub (Nov 14, 2024):

Same here.

<!-- gh-comment-id:2475360849 --> @wahyusaputra commented on GitHub (Nov 14, 2024): Same here.
Author
Owner

@DrPiwo commented on GitHub (Nov 30, 2024):

+1

<!-- gh-comment-id:2508887068 --> @DrPiwo commented on GitHub (Nov 30, 2024): +1
Author
Owner

@kkolago commented on GitHub (Nov 30, 2024):

+1

<!-- gh-comment-id:2508920547 --> @kkolago commented on GitHub (Nov 30, 2024): +1
Author
Owner

@Reknine commented on GitHub (Dec 7, 2024):

+1

<!-- gh-comment-id:2525067786 --> @Reknine commented on GitHub (Dec 7, 2024): +1
Author
Owner

@DangerousBerries commented on GitHub (Dec 26, 2024):

+1

<!-- gh-comment-id:2562296199 --> @DangerousBerries commented on GitHub (Dec 26, 2024): +1
Author
Owner

@quanbisen commented on GitHub (Dec 28, 2024):

+1

<!-- gh-comment-id:2564282424 --> @quanbisen commented on GitHub (Dec 28, 2024): +1
Author
Owner

@Satchitananda commented on GitHub (Dec 29, 2024):

+1

<!-- gh-comment-id:2564706122 --> @Satchitananda commented on GitHub (Dec 29, 2024): +1
Author
Owner

@ociubotaru commented on GitHub (Jan 6, 2025):

Hello,

This is not yet available. I will try to add the support in llama.cpp but not sure how much it may take.

JoanFM was the support for jinaai/jina-embeddings-v3 added ?

<!-- gh-comment-id:2572216246 --> @ociubotaru commented on GitHub (Jan 6, 2025): > Hello, > > This is not yet available. I will try to add the support in llama.cpp but not sure how much it may take. [JoanFM](https://github.com/JoanFM) was the support for jinaai/jina-embeddings-v3 added ?
Author
Owner

@JoanFM commented on GitHub (Jan 6, 2025):

hey @ociubotaru,

no it was not added.

<!-- gh-comment-id:2572935018 --> @JoanFM commented on GitHub (Jan 6, 2025): hey @ociubotaru, no it was not added.
Author
Owner

@alainlizotte commented on GitHub (Jan 6, 2025):

Why not ?

<!-- gh-comment-id:2573065610 --> @alainlizotte commented on GitHub (Jan 6, 2025): Why not ?
Author
Owner

@JoanFM commented on GitHub (Jan 6, 2025):

the effort is higher than expected and could not take the time

<!-- gh-comment-id:2573094260 --> @JoanFM commented on GitHub (Jan 6, 2025): the effort is higher than expected and could not take the time
Author
Owner

@Aelentel commented on GitHub (Jan 14, 2025):

the effort is higher than expected and could not take the time

are you still considering working on it ?

<!-- gh-comment-id:2589756161 --> @Aelentel commented on GitHub (Jan 14, 2025): > the effort is higher than expected and could not take the time are you still considering working on it ?
Author
Owner

@AlgorithmicKing737 commented on GitHub (Jan 25, 2025):

any update yet?

<!-- gh-comment-id:2613800263 --> @AlgorithmicKing737 commented on GitHub (Jan 25, 2025): any update yet?
Author
Owner

@ikbal-nayem commented on GitHub (Feb 16, 2025):

Please bring EN-BN support also.

<!-- gh-comment-id:2661471580 --> @ikbal-nayem commented on GitHub (Feb 16, 2025): Please bring EN-BN support also.
Author
Owner

@Ero-Cat commented on GitHub (Feb 26, 2025):

+1

<!-- gh-comment-id:2683754111 --> @Ero-Cat commented on GitHub (Feb 26, 2025): +1
Author
Owner

@jpapenhagen commented on GitHub (Mar 6, 2025):

+1

<!-- gh-comment-id:2703818303 --> @jpapenhagen commented on GitHub (Mar 6, 2025): +1
Author
Owner

@yunfengsay commented on GitHub (Mar 28, 2025):

Waiting for this

<!-- gh-comment-id:2762173840 --> @yunfengsay commented on GitHub (Mar 28, 2025): Waiting for this
Author
Owner

@xylagbx commented on GitHub (Apr 4, 2025):

+1

<!-- gh-comment-id:2777385909 --> @xylagbx commented on GitHub (Apr 4, 2025): +1
Author
Owner

@sebzur commented on GitHub (Apr 21, 2025):

+1

<!-- gh-comment-id:2818505486 --> @sebzur commented on GitHub (Apr 21, 2025): +1
Author
Owner

@Muhammad-Abdullah-Nizami commented on GitHub (Apr 29, 2025):

+1

<!-- gh-comment-id:2838474921 --> @Muhammad-Abdullah-Nizami commented on GitHub (Apr 29, 2025): +1
Author
Owner

@deepak-stratforge commented on GitHub (May 14, 2025):

+1

<!-- gh-comment-id:2879181125 --> @deepak-stratforge commented on GitHub (May 14, 2025): +1
Author
Owner

@alainlizotte commented on GitHub (May 15, 2025):

Is it GGUF ?

<!-- gh-comment-id:2885047561 --> @alainlizotte commented on GitHub (May 15, 2025): Is it GGUF ?
Author
Owner

@tahitimoon commented on GitHub (May 24, 2025):

+1

<!-- gh-comment-id:2906907343 --> @tahitimoon commented on GitHub (May 24, 2025): +1
Author
Owner

@andyvk85 commented on GitHub (May 25, 2025):

+1

<!-- gh-comment-id:2907805640 --> @andyvk85 commented on GitHub (May 25, 2025): +1
Author
Owner
<!-- gh-comment-id:2907821349 --> @alew3 commented on GitHub (May 25, 2025): Doesn't this work? https://docs.vllm.ai/en/stable/getting_started/examples/offline_inference/embed_jina_embeddings_v3.html?h=jina
Author
Owner

@Chleba commented on GitHub (Jul 2, 2025):

There is already new version 4:
https://huggingface.co/jinaai/jina-embeddings-v4

Could we please have a official support for this embedding model ?

Thanks 👍

<!-- gh-comment-id:3027900795 --> @Chleba commented on GitHub (Jul 2, 2025): There is already new version 4: https://huggingface.co/jinaai/jina-embeddings-v4 Could we please have a official support for this embedding model ? Thanks 👍
Author
Owner

@bkaraujo commented on GitHub (Sep 30, 2025):

+1

<!-- gh-comment-id:3353525166 --> @bkaraujo commented on GitHub (Sep 30, 2025): +1
Author
Owner

@AlhenaaS commented on GitHub (Nov 4, 2025):

+1

<!-- gh-comment-id:3483391704 --> @AlhenaaS commented on GitHub (Nov 4, 2025): +1
Author
Owner

@proemian commented on GitHub (Nov 17, 2025):

+1

<!-- gh-comment-id:3542788909 --> @proemian commented on GitHub (Nov 17, 2025): +1
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#4381