An error occurred while processing chat request: Error: OpenAI or Azure OpenAI API key not found

[MohamedElKasimi]

I get this error when I enter a query on perplexica. But i dont want it to use OpenAI i want it to use ollama. I added the url for ollama on my local machine so why is it looking for OpenAI api key?

[delfireinoso]

Ollama has a special address from
Docker

You should update your config before pull docker

[MohamedElKasimi]

[GENERAL]
SIMILARITY_MEASURE = "cosine" # "cosine" or "dot"
KEEP_ALIVE = "5m" # How long to keep Ollama models loaded into memory. (Instead of using -1 use "-1m")

[MODELS.OPENAI]
API_KEY = ""

[MODELS.GROQ]
API_KEY = ""

[MODELS.ANTHROPIC]
API_KEY = ""

[MODELS.GEMINI]
API_KEY = ""

[MODELS.CUSTOM_OPENAI]
API_KEY = ""
API_URL = ""
MODEL_NAME = ""

[MODELS.OLLAMA]
API_URL ="http://host.docker.internal:11434"

[MODELS.DEEPSEEK]
API_KEY = ""

[MODELS.LM_STUDIO]
API_URL = "" # LM Studio API URL - http://host.docker.internal:1234

[API_ENDPOINTS]
SEARXNG = "" # SearxNG API URL - http://localhost:32768

this is my config.toml file. I changed it, composed dowm, system prune --all and rebuilt it and everything and i still have the same issue.

[delfireinoso]

You need to update for SEARXNG too

After this in the UI (your browser) you need to configure Ollama for the model and a model and choose internal or a suited embedding model

Those models should be installed first with Ollama CLI

[ChvckN0rri5]

Ollama doesn't require an API key. Instead just put something random for the key like "no-key" and it should work. Ollama will ignore the API key if it isn't enabled in Ollama itself.

Try it out and post your results here. It behave the same way with the API key, like any other UI on the web. It's only because the loose OpenAI "standard" requires it.