Refer to Model Configs for how to set the environment variables for your particular deployment.

Note: While we support self hosted LLMs, you will get significantly better responses with a more powerful model like GPT-4.

​
What is Ollama

Ollama provides an easy way to host LLMs locally and to provide a REST API for the model. Refer to the following resources to get started:

Once you start a model with a command like ollama run llama2, you can verify that the API works with a curl request:

curl http://localhost:11434/api/generate -d '{
  "model": "llama2",
  "prompt":"Why is the sky blue?"
}'

​
Set Onyx to use Ollama

On the LLM page in the Admin Panel add a Custom LLM Provider with the following settings:

Note: For the API Base, you should point Onyx to wherever Ollama is running. To point to a Docker container running at http://localhost:11434, use host.docker.internal instead of localhost (as shown in the screenshots).

You may also want to update some of the environment variables depending on your model choice / how you’re running Ollama (e.g. on CPU vs GPU):

# Let's also make some changes to accommodate the weaker locally hosted LLM
QA_TIMEOUT=120  # Set a longer timeout, running models on CPU can be slow
# Always run search, never skip
DISABLE_LLM_CHOOSE_SEARCH=True
# Don't use LLM for reranking, the prompts aren't properly tuned for these models
DISABLE_LLM_CHUNK_FILTER=True
# Don't try to rephrase the user query, the prompts aren't properly tuned for these models
DISABLE_LLM_QUERY_REPHRASE=True
# Don't use LLM to automatically discover time/source filters
DISABLE_LLM_FILTER_EXTRACTION=True
# Uncomment this one if you find that the model is struggling (slow or distracted by too many docs)
# Use only 1 section from the documents and do not require quotes
# QA_PROMPT_OVERRIDE=weak
HuggingFace Inference APIFastChat