# Has inference API stopped returning text embeddings?

**URL:** <https://discuss.huggingface.co/t/has-inference-api-stopped-returning-text-embeddings/155552>\
**Category:** Inference Endpoints on the Hub\
**Created:** [May 16, 2025, 3:02pm UTC](https://discuss.huggingface.co/t/has-inference-api-stopped-returning-text-embeddings/155552 "2025-05-16T15:02:39Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![cassiasamp](https://sea2.discourse-cdn.com/hellohellohello/user_avatar/discuss.huggingface.co/cassiasamp/32/47678_2.png) [@cassiasamp](https://discuss.huggingface.co/u/cassiasamp)\
**Post date:** [May 16, 2025, 3:02pm UTC](https://discuss.huggingface.co/t/has-inference-api-stopped-returning-text-embeddings/155552/1 "2025-05-16T15:02:39Z")

</div>

Hi everyone, I noticed there was a change in the inference API (unfortunately, without any previous warning) and I wanted to know.. will it stop returning text embeddings?

It seems it is mainly returning text similarity now. I’m wondering if there is really a change, i.e. if it will happen for all models, and if I can keep obtaining the embeddings in some other way.

To give more information, we have an application that uses the formerly returned embeddings, we were using it and noticed that for some models they weren’t being returned anymore. Now, we are trying to figure out what to do and to find more information about this.

Ps: it would be awesome if HF made us aware about API changes prior to them happening.

---

<div class="post-metadata">

**Author:** ![John6666](https://sea2.discourse-cdn.com/hellohellohello/user_avatar/discuss.huggingface.co/john6666/32/27664_2.png) [@John6666](https://discuss.huggingface.co/u/John6666)\
**Post date:** [May 17, 2025, 4:34am UTC](https://discuss.huggingface.co/t/has-inference-api-stopped-returning-text-embeddings/155552/2 "2025-05-17T04:34:22Z")

</div>

I don’t see anything particularly noteworthy in the TGI release notes…  
I think any changes to the return values will depend on this library.

The following changes appear to have been made to the embedding model.

> **[sentence-transformers/all-MiniLM-L6-v2 · Updated feature-extraction API URL](https://huggingface.co/sentence-transformers/all-MiniLM-L6-v2/discussions/116)**
>
> Hello!

> **[Releases · huggingface/text-generation-inference](https://github.com/huggingface/text-generation-inference/releases)**
>
> Large Language Model Text Generation Inference. Contribute to huggingface/text-generation-inference development by creating an account on GitHub.