Quick answer
Install the Ollama app from ollama.com, open Terminal and type ollama run inkling-2-mini. Ollama downloads the model (about 20GB) and opens a chat right in Terminal. You need a Mac with Apple silicon and at least 32GB of memory, plus around 25GB of free disk space. The first download takes a while; after that it starts in seconds and works offline.
Key takeaways
- 32GB of memory or more is the stated minimum for comfortable use.
- One command downloads and starts the model.
- Test with a real task before relying on it.
- Remove it in one command if you need the space back.
📘 Main article: This guide follows our Inkling 2 Mini Review: Can a 26B Local AI Model Beat Qwen 3.8 27B?, which explains what the model is good at and where it falls short.
Inkling 2 Mini from Thinking Machines is one of the strongest AI models you can run on your own computer. Ollama is a free app that makes downloading and running models like this simple. This guide uses the official Ollama model page and takes about 20 minutes, most of it waiting for the download.
What do you need before you start?
| Requirement | What you need | How to check |
|---|---|---|
| Computer | A Mac with Apple silicon (M-series chip) | Apple menu, About This Mac, “Chip” |
| Memory | 32GB or more | Same window, “Memory” |
| Disk space | About 25GB free | System Settings, General, Storage |
| Internet | For the first download only | About 20GB to download |
⚠️ Warning: With 16GB of memory or less, the model won’t run well, if at all. Use a cloud AI tool or a much smaller model instead. Windows and Linux PCs with a powerful NVIDIA graphics card may also work; check the Ollama model page for current support.
Not sure what your computer can handle? Read How Much Memory Do You Need to Run an Open-Weight AI Model?.
How do you install Ollama?
- Go to ollama.com/download and download the Mac app.
- Open the downloaded file and drag Ollama into your Applications folder.
- Open Ollama. A small llama icon appears in your menu bar, which means it is running.
- If you already had Ollama, update it to the latest version, as new models often need it.
How do you download and run Inkling 2 Mini?
- Open the Terminal app (press Command + Space and type “Terminal”).
- Type
ollama run inkling-2-miniand press Return. - Wait for the download to finish. You’ll see a progress bar for about 20GB.
- When you see a prompt that says “Send a message”, type a question and press Return.
- To leave the chat, type
/bye.
💡 Tip: If you prefer not to use Terminal, the Ollama app has its own chat window. Open it from the menu bar icon and choose inkling-2-mini from the model list once it has downloaded.
How do you check it is working well?
- Ask something simple, such as “Summarise what a looped transformer is in three sentences.” It should answer quickly.
- Give it a real task from your work, such as tidying a list of product names or writing a small spreadsheet formula.
- Check the answer, especially facts. Small models know fewer facts; see How to Ask AI for Answers You Can Check in Minutes.
- Type
ollama psin a new Terminal window to see that the model is loaded and how much memory it uses.
Can you use it with other apps?
Yes. Ollama runs a local server that other apps can connect to. The model page lists ready-made commands for coding tools, for example:
ollama launch claude --model inkling-2-minito use it inside Claude Code.ollama launch opencode --model inkling-2-minifor OpenCode.
Many desktop chat apps also let you choose an Ollama model in their settings. Look for “Ollama” or “local model”.
What if it’s slow or won’t run?
- Very slow answers: close other heavy apps, especially browsers with lots of tabs. If your Mac has exactly 32GB, memory is tight.
- “Model not found” or an error: update Ollama and check the spelling:
inkling-2-mini. - Download stops: run the same command again; it carries on where it left off.
- Need the space back: type
ollama rm inkling-2-minito delete the model.
Your next step: check your Mac’s memory now. If it has 32GB or more, install Ollama and start the download so it’s ready when you need it.
Frequently asked questions
Is Ollama free?
Yes. Ollama is free to download and use for running models on your own computer.
How big is Inkling 2 Mini?
About 20GB on disk. You need a Mac with at least 32GB of memory to run it comfortably.
Does it work without the internet?
Yes. After the first download, the model runs entirely on your computer.
Can I run Inkling 2 Mini on Windows?
Possibly, with a powerful NVIDIA graphics card. Check the Ollama model page for current support.
The bottom line
If your Mac has the memory, running Inkling 2 Mini is a one-command job. Install Ollama, run the model, test it on a real task and check the results. Then decide whether it earns a place in your workflow.
Related posts
- Inkling 2 Mini Review: Can a 26B Local AI Model Beat Qwen 3.8 27B?
- How Much Memory Do You Need to Run an Open-Weight AI Model?
- Mac, NVIDIA or Your Old Computer: Which Should Run Your AI Agents?
