# How to use weaviate with LM Studio?
**URL:** https://forum.weaviate.io/t/how-to-use-weaviate-with-lm-studio/3136
**Category:** Support
**Created:** [July 23, 2024, 7:56pm UTC](https://forum.weaviate.io/t/how-to-use-weaviate-with-lm-studio/3136 "2024-07-23T19:56:26Z")
**Posts on this page:** 4
**Page:** 1
### Author:  [@ctindel](https://forum.weaviate.io/u/ctindel)
#### Post date: [July 23, 2024, 7:56pm UTC](https://forum.weaviate.io/t/how-to-use-weaviate-with-lm-studio/3136/1 "2024-07-23T19:56:26Z")
### Description
I have one of the weaviate docker configurations running on my laptop and I also have LM Studio running on my laptop so I can serve up the Llama 3 LLM with a local OpenAI compatible endpoint.
However, unlike the ollama module which takes an apiEndpoint parameter, I don’t see how to set a custom local apiEndpoint for the OpenAI module.
What’s the best way to call into an LM Studio LLM from a weaviate module?
---
### Author:  [@DudaNogueira](https://forum.weaviate.io/u/DudaNogueira)
#### Post date: [July 23, 2024, 9:05pm UTC](https://forum.weaviate.io/t/how-to-use-weaviate-with-lm-studio/3136/2 "2024-07-23T21:05:54Z")
hi @ctindel !!
Welcome to our community 🤗
I have not played with LM Studio! Just downloaded it here. Awesome project!
You can set a [base url at query time](https://weaviate.io/developers/weaviate/modules/retriever-vectorizer-modules/text2vec-openai#query-time-parameters) or define it at collection creation.
Here is the code I crafted, based on our quickstart:
```python
import weaviate
from weaviate import classes as wvc
client = weaviate.connect_to_local()
client.collections.delete("Question")
questions = client.collections.create(
name="Question",
vectorizer_config=wvc.config.Configure.Vectorizer.text2vec_openai(
base_url="http://host.docker.internal:1234"
), # If set to "none" you must always provide vectors yourself. Could be any other "text2vec-*" also.
generative_config=wvc.config.Configure.Generative.openai(
base_url="http://host.docker.internal:1234"
) # Ensure the `generative-openai` module is used for generative queries
)
# insert data
import requests, json
resp = requests.get('https://raw.githubusercontent.com/weaviate-tutorials/quickstart/main/data/jeopardy_tiny.json')
data = json.loads(resp.text) # Load data
question_objs = list()
for i, d in enumerate(data):
question_objs.append({
"answer": d["Answer"],
"question": d["Question"],
"category": d["Category"],
})
questions = client.collections.get("Question")
questions.data.insert_many(question_objs)
# now query
questions = client.collections.get("Question")
response = questions.query.near_text(
query="biology",
limit=2
)
print(response.objects[0].properties)
# now let's generate some content
generate = questions.generate.near_text(limit=2, query="biology", grouped_task="generate a tweet about the questions {question}")
```
the output:
> Here’s a tweet about the questions:
>
> “Did you know? Watson & Crick built a model of DNA in 1953! And, did you know that our liver is responsible for removing excess glucose from the blood and storing it as glycogen? Mind blown! #ScienceFacts #DNA #LiverFunction”
>
> ### Explanation:
>
> The response generates a tweet by combining information from the provided questions. The first question asks about the molecular structure of DNA, which Watson & Crick built in 1953. The second question talks about the liver’s role in removing excess glucose and storing it as glycogen. The response combines these facts into a concise and engaging tweet that includes relevant hashtags to make it discoverable by others interested in science and health-related topics.
Looking at LM logs, I can see data going thru it:
and here the generate part:
Let me know if this helps!
Thanks!
---
### Author:  [@ctindel](https://forum.weaviate.io/u/ctindel)
#### Post date: [July 24, 2024, 5:06am UTC](https://forum.weaviate.io/t/how-to-use-weaviate-with-lm-studio/3136/3 "2024-07-24T05:06:59Z")
Yes that is very helpful!
It’s confusing because the docs pages for the text2vec\_openai module ([text2vec-openai | Weaviate - Vector Database](https://weaviate.io/developers/weaviate/modules/retriever-vectorizer-modules/text2vec-openai)) refer to it in camel case like “baseURL”
However now I’m getting this error:
Failed to import 1 objects
e.g. Failed to import object with error: WeaviateInsertManyAllFailedError(‘Every object failed during insertion. Here is the set of all errors: send POST request: Post “[http://host.docker.internal:1234/v1/embeddings](http://host.docker.internal:1234/v1/embeddings)”: context deadline exceeded (Client.Timeout exceeded while awaiting headers)’)
And I’m wondering if generating the vector is just taking too long because its a large object. Is there an equivalent in openai\_text2vec of the text\_fields parameter used in multi2vec-bind ? Like, without using named indexes how do I tell openai\_text2vec which properties to use for generating the vector?
---
### Author:  [@DudaNogueira](https://forum.weaviate.io/u/DudaNogueira)
#### Post date: [July 24, 2024, 8:50pm UTC](https://forum.weaviate.io/t/how-to-use-weaviate-with-lm-studio/3136/4 "2024-07-24T20:50:50Z")
Ok! We had an opportunity to tackle this in our [office hours](https://weaviate.io/community/events), and the issue was that the embedding model was not defined in LM Studio.
When using LM Studio, make sure to also load an embbeding model. So this should work:
```shell
curl http://localhost:1234/v1/embeddings \
-H "Content-Type: application/json" \
-H "Authorization: Bearer THIS_WILL_BE_IGNORED" \
-d '{
"input": "Your text string goes here",
"model": "THIS WILL ALSO BE IGNORED"
}'
```
THanks!