Ollama run returns authentication error on in Windows.
I already received authorization, created a separate access token with the necessary permissions, added Ollama's public SSH key to my account, logged in via huggingface-cli login with the correct access token, then ran the following:
ollama run hf.co/google/gemma-3-27b-it-qat-q4_0-gguf
pulling manifest
Error: pull model manifest: 401: {"error":"Invalid username or password."}
I have no idea what I could be doing wrong here and I would like some assistance. I tried downloading the model separately and importing it from gguf via Ollama and that didn't seem to work neither. Is there an alternative way to get access to this model? Because I feel like I'm chasing my own tail here.
I had the same error
In LMStudio, the model is visible in the search, but when trying to download it gives an error - access denied
I needed to upload the public key for Ollama and finally found it located at:
cat /usr/share/ollama/.ollama/id_ed25519.pub
Thought I'd share this in case anyone else is looking for it!
Try this for Windows
https://github.com/ollama/ollama/issues/4901
