[GH-ISSUE #4491] Pulling using API - Session timeout (5 minutes) #2810

Closed
opened 2026-04-12 13:08:36 -05:00 by GiteaMirror · 1 comment
Owner

Originally created by @pelletier197 on GitHub (May 17, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/4491

What is the issue?

When using the REST API to pull models, the PULL request seems to timeout for large models (llama3).

This is linked to this issue.

Is there any way to override the default session timeout when pulling models ? I noticed that the Generate endpoint accepts a KeepAlive parameter to override this. Could this be possible for pull as well ?

OS

Linux

GPU

Nvidia

CPU

Intel

Ollama version

01.38

Originally created by @pelletier197 on GitHub (May 17, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/4491 ### What is the issue? When using the REST API to pull models, the `PULL` request seems to timeout for large models (llama3). This is linked to [this issue](https://github.com/ollama/ollama-js/issues/72). Is there any way to override the default session timeout when pulling models ? I noticed that the `Generate` endpoint accepts a `KeepAlive` [parameter](https://github.com/ollama/ollama/blob/7e1e0086e7d18c943ff403a7ca5c2d9ce39f3f4b/server/routes.go#L136) to override this. Could this be possible for `pull` as well ? ### OS Linux ### GPU Nvidia ### CPU Intel ### Ollama version 01.38
GiteaMirror added the bug label 2026-04-12 13:08:36 -05:00
Author
Owner

@dhiltgen commented on GitHub (Jul 25, 2024):

I'm fairly confident this is a client-side issue. Enable streaming, or set a longer client-side timeout

<!-- gh-comment-id:2251518535 --> @dhiltgen commented on GitHub (Jul 25, 2024): I'm fairly confident this is a client-side issue. Enable [streaming](https://github.com/ollama/ollama/blob/main/docs/api.md#pull-a-model), or set a longer client-side timeout
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#2810