github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-07 08:30:05 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #13381] Custom cloud models with remote server configuration #70896

New Issue
Open
opened 2026-05-04 23:24:58 -05:00 by GiteaMirror · 3 comments
GiteaMirror commented 2026-05-04 23:24:58 -05:00
Owner
Copy Link

Originally created by @uvulpos on GitHub (Dec 8, 2025).
Original GitHub issue: https://github.com/ollama/ollama/issues/13381

Hey there,

I want to have the opportunity to proxy traffic between two ollama instances so I can share one big graphics card, or work with less power mobile, without to reconfigure my application for development. Is there a possibility to allow that with the cloud models?

Conversation Summary:

  • Possibility to connect ollama to ollama instance via custom cloud model (proxy)
  • Connection configuration inside the cloud model
    • Authentication (Bearer, Basic,...)
Originally created by @uvulpos on GitHub (Dec 8, 2025). Original GitHub issue: https://github.com/ollama/ollama/issues/13381 Hey there, I want to have the opportunity to proxy traffic between two ollama instances so I can share one big graphics card, or work with less power mobile, without to reconfigure my application for development. Is there a possibility to allow that with the cloud models? --- ## Conversation Summary: - Possibility to connect ollama to ollama instance via custom cloud model (proxy) - Connection configuration inside the cloud model - Authentication (Bearer, Basic,...)
GiteaMirror added the feature request label 2026-05-04 23:24:58 -05:00
GiteaMirror commented 2026-05-04 23:24:59 -05:00
Author
Owner
Copy Link

@fcorneli commented on GitHub (Dec 8, 2025):

With possibility to add Bearer or Basic authentication headers.

<!-- gh-comment-id:3628917177 --> @fcorneli commented on GitHub (Dec 8, 2025): With possibility to add Bearer or Basic authentication headers.
GiteaMirror commented 2026-05-04 23:24:59 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Dec 8, 2025):

Some clarity on the goal would help with coming up with a solution.

  1. You want multiple ollama servers to use a one big graphics card: run multiple servers, they will share the card.
  2. You want to run multiple servers that use a cloud model: run multiple servers, they will share the cloud.
  3. You want to run your own models in Ollama Cloud. Not supported at this time.
  4. You want to run an ollama server offsite, and have multiple clients connect: run nginx/caddy to provide access control.
<!-- gh-comment-id:3628964538 --> @rick-github commented on GitHub (Dec 8, 2025): Some clarity on the goal would help with coming up with a solution. 1. You want multiple ollama servers to use a one big graphics card: run multiple servers, they will share the card. 2. You want to run multiple servers that use a cloud model: run multiple servers, they will share the cloud. 3. You want to run your own models in Ollama Cloud. Not supported at this time. 4. You want to run an ollama server offsite, and have multiple clients connect: run nginx/caddy to provide access control.
GiteaMirror commented 2026-05-04 23:25:00 -05:00
Author
Owner
Copy Link

@uvulpos commented on GitHub (Dec 8, 2025):

The idea is to share a big ollama gpu server for many devs, but in a confidential environment to save resources (like ollama cloud but self-hosted). You can create own cloud models that contain the configuration of the remote server and maybe even alternative models, in case the bigger model is not available due to no cell or something. So you can configure your ollama instance and it routes the request automatically and even stores authentication via keychain vault, windows vault etc.
.
Image
.
The usage could be guided to create that model, you can push it to a private company registry and after that you have to insert the bearer token once to authenticate (if required) and then just proxy all requests to the big gpu servers or internally if not available. I have a docker alike usage in mind.
.
Image
.
Having an ollama cluster sounds also interesting, but that's another issue

<!-- gh-comment-id:3629152003 --> @uvulpos commented on GitHub (Dec 8, 2025): The idea is to share a big ollama gpu server for many devs, but in a confidential environment to save resources (like ollama cloud but self-hosted). You can create own cloud models that contain the configuration of the remote server and maybe even alternative models, in case the bigger model is not available due to no cell or something. So you can configure your ollama instance and it routes the request automatically and even stores authentication via keychain vault, windows vault etc. . <img width="1016" height="596" alt="Image" src="https://github.com/user-attachments/assets/4a53b9a3-43f6-4a11-93b2-316f001119f6" /> . The usage could be guided to create that model, you can push it to a private company registry and after that you have to insert the bearer token once to authenticate (if required) and then just proxy all requests to the big gpu servers or internally if not available. I have a docker alike usage in mind. . <img width="804" height="413" alt="Image" src="https://github.com/user-attachments/assets/03483e02-aebc-4a31-9a7f-cfbdf5fa2a0c" /> . Having an ollama cluster sounds also interesting, but that's another issue
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
amd
api
app
bug
build
cli
cloud
compatibility
context-length
create
docker
documentation
embeddings
feature request
feedback wanted
good first issue
gpt-oss
gpu
harmony
help wanted
image
install
intel
js
launch
linux
macos
memory
mlx
model
needs more info
networking
nvidia
ollama.com
performance
pull-request
Mirrored from GitHub Pull Request
 python
question
registry
rendering
thinking
tools
top
vulkan
windows
wsl
No Label feature request
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
GiteaMirror ninjasurge
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#70896
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?