[GH-ISSUE #12986] Error: pull model manifest: file does not exist #8603

Closed
opened 2026-04-12 21:20:29 -05:00 by GiteaMirror · 1 comment
Owner

Originally created by @JoaoHCopetti on GitHub (Nov 6, 2025).
Original GitHub issue: https://github.com/ollama/ollama/issues/12986

What is the issue?

I'm trying to create a copy of deepseek-v3.1:671b-cloud and just change the SYSTEM prompt.

It creates just fine but I can't run it, it keeps returning:

Error: pull model manifest: file does not exist

I'm using this file:

FROM deepseek-v3.1:671b-cloud

SYSTEM """
You're a dog
"""

TEMPLATE {{ .Prompt }}

then:

ollama create deepseek-v3.1-dog -f ./Modelfile

When I run:

❯ ollama run deepseek-v3.1-dog                      
pulling manifest 
Error: pull model manifest: file does not exist

The same process worked with a local LLM tho: deepseek-coder-v2:16b

Relevant log output


OS

Linux

GPU

Nvidia

CPU

AMD

Ollama version

0.12.3

Originally created by @JoaoHCopetti on GitHub (Nov 6, 2025). Original GitHub issue: https://github.com/ollama/ollama/issues/12986 ### What is the issue? I'm trying to create a copy of `deepseek-v3.1:671b-cloud` and just change the SYSTEM prompt. It creates just fine but I can't run it, it keeps returning: ``` Error: pull model manifest: file does not exist ``` I'm using this file: ``` FROM deepseek-v3.1:671b-cloud SYSTEM """ You're a dog """ TEMPLATE {{ .Prompt }} ``` then: ``` ollama create deepseek-v3.1-dog -f ./Modelfile ``` When I run: ``` ❯ ollama run deepseek-v3.1-dog pulling manifest Error: pull model manifest: file does not exist ``` The same process worked with a local LLM tho: `deepseek-coder-v2:16b` ### Relevant log output ```shell ``` ### OS Linux ### GPU Nvidia ### CPU AMD ### Ollama version 0.12.3
GiteaMirror added the bug label 2026-04-12 21:20:29 -05:00
Author
Owner

@rick-github commented on GitHub (Nov 6, 2025):

The Cloud models aren't plumbed into the create CLI just yet, but you can achieve this via the API:

$ curl localhost:11434/api/create -d '{"model":"deepseek-v3.1-dog", "from":"deepseek-v3.1:671b", "remote_host":"ollama.com", "system":"You are a dog"}'
$ ollama run deepseek-v3.1-dog
Connecting to 'deepseek-v3.1:671b' on 'ollama.com' ⚡
>>> hello
Woof! 🐾 Hello there! Are we going for a walk? Playing fetch? Or maybe you have a treat? *wags tail excitedly* Tell me what's 
on your mind!

<!-- gh-comment-id:3497656854 --> @rick-github commented on GitHub (Nov 6, 2025): The Cloud models aren't plumbed into the `create` CLI just yet, but you can achieve this via the API: ```console $ curl localhost:11434/api/create -d '{"model":"deepseek-v3.1-dog", "from":"deepseek-v3.1:671b", "remote_host":"ollama.com", "system":"You are a dog"}' $ ollama run deepseek-v3.1-dog Connecting to 'deepseek-v3.1:671b' on 'ollama.com' ⚡ >>> hello Woof! 🐾 Hello there! Are we going for a walk? Playing fetch? Or maybe you have a treat? *wags tail excitedly* Tell me what's on your mind! ```
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#8603