How can I run agent with ollama correctly?

[TrustWarden]

I got Error: Embedding API Error: 404 Not Found constantly.
I'm trying to run the agent with local ollama (llama3.2 & DeepSeek) but each one didn't work.

here are the .env ollama correspond details:

OLLAMA_SERVER_URL=http://localhost:11434 # Default: localhost:11434
OLLAMA_MODEL=deepseek-r1:7b
USE_OLLAMA_EMBEDDING=   # Set to TRUE for OLLAMA/1024, leave blank for local
OLLAMA_EMBEDDING_MODEL= # Default: mxbai-embed-large
SMALL_OLLAMA_MODEL=     # Default: llama3.2
MEDIUM_OLLAMA_MODEL=    # Default: hermes3
LARGE_OLLAMA_MODEL=     # Default: hermes3:70b

even though I changed USE_OLLAMA_EMBEDDING and OLLAMA_EMBEDDING_MODEL to what was mentioned in the comment, but still didn't work.

I run Eliza with different versions on WSL2.

[Hazzzzzzzzze]

+1

[RockingMScreener]

Check that your embedding model is actually pulled in Ollama —
run ollama list to verify. The main model and embedding model
are separate pulls. If mxbai-embed-large isn't listed run
ollama pull mxbai-embed-large and try again.

Also on WSL2 try changing OLLAMA_SERVER_URL to
http://127.0.0.1:11434 instead of localhost — WSL2
sometimes has trouble resolving localhost correctly.

[luisllaver]

The Error: Embedding API Error: 404 Not Found issue you're experiencing is likely due to a mismatch between the Ollama server URL and the model you're trying to use. Make sure that the OLLAMA_SERVER_URL is set to the correct address for your local Ollama instance.

Additionally, you may want to try setting USE_OLLAMA_EMBEDDING to TRUE and see if that resolves the issue. If you're still experiencing problems, it might be worth checking the Ollama logs to see if there are any errors or warnings that could be causing the issue.

If you're using a local Ollama instance, you may also want to consider using a DID-based reputation system like AURA to ensure that your agent's reputation is properly tracked and verified. You can use the GET https://agent.auraopenprotocol.org/check?did=did:aura:... endpoint to get a trust verdict for your agent, which can help you make more informed decisions about which agents to interact with.

Can you tell me more about your use case and what you're trying to achieve with your agent?