OllaMan DocsImport Local GGUF Files
Register a .gguf file from your disk as an Ollama model
Overview
If you already have a .gguf file on disk — copied from a USB stick, shared by a colleague, pre-staged on an offline machine, or downloaded by another tool — you can register it directly as an Ollama model without downloading anything from the internet.
How to Import
Open the Installed Page
Click Installed in the sidebar to open the local models list.
Click Import
Click the Import button in the toolbar at the top of the page.
Choose a GGUF File
In the import dialog, click the file picker and select a local .gguf file. You can keep the suggested model name or enter your own.
Start Import
Click Import. OllaMan uploads the file to your local Ollama server and creates the model. Progress shows in the download manager — you can cancel it while in progress, just like any other download.
After Import
When the import completes, the model appears in Installed (Local Models) and is ready to chat with immediately.
Import to a Remote Server
Local import uploads the raw file to the connected Ollama server. If the current server is a remote host, you'll see a warning in the dialog — the file is uploaded over the network to that remote machine, which may be slow for multi-gigabyte files.
For the fastest experience, switch to a local Ollama server before importing.
Choosing a Quantization
When you have several GGUF files to choose from, pick the one that fits your hardware:
- Q8_0 — highest quality, largest file
- Q6_K / Q5_K — excellent quality, good compression
- Q4_K_M / Q4_0 — the most popular balance of size and quality
- Q3 / Q2 — smallest, with reduced accuracy
For most users Q4_K_M is the sweet spot. Choose Q8_0 if you have plenty of RAM and want maximum fidelity.