Bonsai 2: Prism ML's ternary vision models land on Ollama in four sizes
Prism ML has released Bonsai 2, a model family built on ternary weights (−1, 0, +1). According to its description, the models are about half the size of standard 4-bit models at near-full quality, and they're available on Ollama in four sizes.
What it's about
Prism ML has released Bonsai 2, a family of vision models available in the Ollama library. What sets it apart: the models are built for ternary weights. Weights are the numbers a model learns during training; ternary means each weight can take only three values: −1, 0 or +1. According to the description, this makes the models about half the size of typical 4-bit models, with quality close to full precision. That's a claim from the vendor. The source doesn't cite any benchmarks.
Sizes and capabilities
Bonsai 2 comes in four sizes: 2B, 4B, 9B and 27B (billions of parameters). All sizes support:
- Thinking: the model reasons step by step before it answers.
- Tool calling: the model can call functions and external tools.
- Image input: images can be processed together with text.
- 262K-token context window: roughly 262,000 tokens (text units) can sit in the model's working memory at once.
Using it with Ollama
The model starts with ollama run bonsai-2. Ollama also offers ollama launch to plug Bonsai 2 straight into coding and agent tools: Claude (ollama launch claude --model bonsai-2), opencode, hermes and openclaw. The page lists about 43,400 downloads and was last updated six days ago.
What you need to try it
The source gives no concrete hardware requirements. As a rough rule of thumb, standard 4-bit models need about 0.6 GB per billion parameters, and the vendor says ternary models need roughly half that. That suggests the 2B and 4B variants should run on ordinary laptops or near-smartphone-class devices, 9B on GPUs with around 4–6 GB, and 27B more likely with 10–12 GB (the Ollama downloads are about 3.2 GB for 9B and 8.3 GB for 27B). These are estimates. A long context costs extra memory. Macs and CPU-only setups work with Ollama in principle, but they're slower.
Our take
Ternary weights are one way to compress models heavily without losing much quality. That matters for local users because larger models could run on smaller hardware. On top of that, all sizes offer vision, tool use and long context. What remains open is how big the quality loss really is: the source provides no evidence such as benchmarks, base model or license. The "close to full precision" claim comes from the vendor, so check it yourself.
Sources
- Ollama-Bibliothek ollama.com
Try it yourself: Bonsai 2
Explained step by step, no prior knowledge needed – on your own computer or free in your browser.
What can I do with it?
Bonsai 2 understands text and images, can think before answering and can call tools. Thanks to its very compact ternary weights, it runs with less memory than typical models.
Who is it worth it for?
Interesting for anyone who wants to test vision models with long context on limited hardware. If you need reliable benchmarks, run your own comparison first.
What your computer needs
Which variant is right for me?
Many models come in several sizes or "quantizations." These are compressed versions: smaller and faster, but slightly less accurate. Rule of thumb: pick the largest variant whose memory requirement fits in your VRAM (video memory). The scripts below choose this for you automatically.
| Variant | Download | Needs approx. | Compatible graphics cards |
|---|---|---|---|
2b |
1.6 GB | 3 GB | almost any graphics card with 4 GB or more (e.g. GTX 1650, RTX 3050) |
9b Default |
3.2 GB | 5 GB | e.g. RTX 2060, RTX 3050 8 GB, RX 6600 |
27b |
8.3 GB | 10 GB | e.g. RTX 3060 12 GB, RTX 4070, RTX 5070, RX 7700 XT |
Start with 2B or 4B on weak hardware, 9B for mid-range GPUs, and 27B if you have plenty of memory. The source gives no specifics; this is a general recommendation.
- According to the vendor, about half the size of standard 4-bit models.
- The long context (262K tokens) costs extra memory; start with shorter contexts.
- The 27B variant needs significantly more memory than 2B to 9B.
- Running on a Mac or CPU is possible with Ollama, but slower.
How do I find out what my computer has?
- Windows: Press Ctrl + Shift + Esc → Task Manager → Performance tab → click GPU on the left. At the bottom you'll see Dedicated GPU memory (that's your VRAM). You'll find system memory under Memory.
- Mac: Apple menu at the top left → About This Mac. You'll see your chip (e.g. M2) and memory (e.g. 16 GB). On Apple chips, graphics and processor share this memory.
- Linux: Type
nvidia-smiin the terminal (shows your graphics card and its memory) andfree -hfor system memory.
Guide
Google Colab (in your browser, no installation)
Google Colab is a free service from Google. It gives you a computer with a graphics card in your browser for a few hours. All you need is a Google account. For this model: runs for free – The free Tesla T4 GPU with 15 GB is enough.
- Download the notebook. Click the button and save the file.
- Open Colab. Go to colab.research.google.com and sign in with your Google account.
- Upload the notebook. In the window that appears, click Upload on the left (or choose File → Upload notebook at the top) and select the file
colab.ipynb. - Turn on the GPU. In the top menu, go to Runtime → Change runtime type → under Hardware accelerator, select T4 GPU → Save.
- Start everything. Runtime → Run all (keyboard shortcut Ctrl + F9). If you see the warning “This notebook was not authored by Google,” click Run anyway.
- Wait. The cells run one after another. A green checkmark appears to the left of each cell when it's done. Installation and download take about 5 minutes here.
- Try it out. The result appears below the last cell. To test something else, change the text in the cell and run it again with Shift + Enter. The last cell starts a proper chat with an input field.
- Clean up. When you're done: Runtime → Disconnect and delete runtime. This saves your free quota.
Good to know: The free GPU is limited (usually a few hours per day). After about 90 minutes of inactivity, Colab disconnects, and you'll have to start everything over.
Windows 10 / 11
The easiest way is Ollama, a free program that launches AI models with a single command.
- Update your graphics driver (NVIDIA cards only): Install the NVIDIA app or the latest driver. For AMD cards, use the latest Adrenalin driver.
- Install Ollama. Download the file OllamaSetup.exe from ollama.com/download/windows, double-click it, and choose Install. Afterward, a small llama icon appears in the taskbar at the bottom right.
- Open a terminal window. Press the Windows key, type PowerShell and press Enter. A blue or black window opens.
- Start the model. Copy this command ("Copy" button), paste it into the window with a right-click, and press Enter:
The first time, Ollama downloads the model (3.2 GB (about 4 min at 100 Mbit/s)). After that, it's instant.ollama run bonsai-2:9b - Start typing. Once
>>> Send a messageappears, type your question and press Enter. Type/byeto end the chat.
Not enough memory? Pick a smaller variant from the table above, e.g. ollama run bonsai-2:2b.
Prefer automatic? A script that picks the right variant for you
- Download both files into the same folder: try_windows.ps1 demo.py
- Open the folder in File Explorer, right-click an empty spot and choose Open in Terminal.
- Paste this command and press Enter:
powershell -ExecutionPolicy Bypass -File try_windows.ps1
If needed, the script installs Ollama via winget, measures your VRAM (video memory) and downloads the matching variant.
macOS
Macs with an Apple chip (M1 to M5) are ideal. Older Intel Macs are very slow.
- Install Ollama. Download it from ollama.com/download/mac, open the file, and drag Ollama into your Applications folder. Launch Ollama once – a llama appears in the menu bar at the top.
- Open Terminal. Press Cmd + Space, type Terminal, and press Enter.
- Start the model. Paste the command (Cmd + V) and press Enter:
The first time, the model is downloaded (3.2 GB (about 4 min at 100 Mbit/s)).ollama run bonsai-2:9b - Start typing. Type your question and press Enter. Quit with
/bye.
Linux
- Check your GPU.
nvidia-smishould show your card. If it doesn't, install the driver on Ubuntu and restart:sudo ubuntu-drivers install - Install Ollama:
curl -fsSL https://ollama.com/install.sh | sh - Start the model:
Exit withollama run bonsai-2:9b/bye. Useollama psto check whether the model is running on the GPU. - Or let it pick the right variant automatically:
bash try_linux.sh
Try it
Describe what you see in this image and summarize it in three sentences.Explain step by step how to make a sourdough bread dough.Write a Python function that reads a CSV file and calculates the sum of one column.If something goes wrong
Without a recent enough Ollama version, the model won't start; update Ollama.
Specify the size, for example `bonsai-2:4b`, if the default size is too large (check the exact tags in the Ollama library).
Very long contexts can exhaust your memory; reduce the context length.
“out of memory,” a crash, or your computer freezes
The model is too big for your VRAM (video memory). Pick a smaller variant from the table above and close other programs, especially games and browsers with lots of tabs.
It runs, but extremely slowly
Your CPU is probably doing the work instead of your graphics card. In Ollama, the PROCESSOR column of ollama ps shows how much runs on the GPU. If it says "100% CPU", update your graphics driver or pick a smaller variant that fits entirely into VRAM (video memory).
“ollama” or “python” is not found
Close the terminal window and open it again so it recognizes the newly installed program. For Python on Windows: reinstall it and check Add python.exe to PATH.
Windows: "Running scripts is disabled on this system"
Run the script with exactly the command above: powershell -ExecutionPolicy Bypass -File try_windows.ps1. This allows execution for this one time only.
Error message when loading the model (version, "trust_remote_code", unknown architecture)
Brand-new models often need the latest software version. With Ollama: update Ollama. With Python: pip install -U transformers in the activated environment. The Model Card usually lists the exact version.
Tokenstrom generates the scripts and notebook automatically. Please take a quick look before running them – they're plain, readable text. Mind the model's license: apache-2.0.
Ask the AI assistant
Questions about this article, installation or background? Our assistant knows this page and runs on our own machine, not in someone else's cloud.
The Assistant is an AI (Gemma 4) and can make mistakes – please check important details against the source. Questions are used only to generate an answer and are deleted after one hour. Please don't enter any personal data.
Comments
What do you think? Already tried it? Write it here, no sign-up needed.
Comments are public. Links aren't allowed, and we delete abusive posts. You can delete your own comment at any time in this browser. Learn more in the Privacy Policy.