TokForge Bonsai 2 27B needs about 9.0 GB of memory at its main file (a 6.4 GB file) with an 8K context window, so it fits on a 12 GB GPU. This repository has 3 quants, from 5.9 GB to 7.2 GB. Longer context needs more: at 32K tokens the KV cache alone adds about 8.6 GB.
TokForge Bonsai 2 27B – ternary 27B‑class model for on‑device inference
TokForge Bonsai 2 27B is a 27B‑parameter ternary model based on Qwen3.8‑27B. It offers 262K‑token context and is packaged for llama.cpp and MLX. The model retains 98.2% of FP16 intelligence while being ~5.9 GB in size.
Good at
- Large‑context reasoning
- Low‑bit on‑device inference
- High‑precision retention
Limits
- No vision support
- Requires CUDA/Metal or MLX
- No mixture‑of‑experts
Best for
- Mobile LLM apps
- Edge inference
- Low‑memory deployments
Who it’s for. Developers building on‑device language applications who need a 27B‑class model that fits in a few gigabytes. The model is ideal for Android and Apple Silicon environments with GPU or CPU support.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| — | 5.9 GB | 8.6 GB | Too big | Get file → | |
| — | 6.4 GB | 9.0 GB | Too big | Get file → | |
| — | 7.2 GB | 9.9 GB | Too big | Get file → |
For 8 GB VRAM use the PTQ1_0 file (5.95 GB, ~8.6 GB memory at 8 k context). For 16 GB RAM on Snapdragon 8 Elite Gen 5 or 8 Gen 3 use the PQ2_0 file (7.21 GB, ~9.9 GB memory). The PTQ1_0 file is smaller and works on CPU if GPU is unavailable.
What GPU do you need for TokForge Bonsai 2 27B?
These figures add the file size, an fp16 KV cache at an 8K context window (worked out from this model’s own header) and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.
- 4 GB VRAM: no quant fits at an 8K context.
- 6 GB VRAM: no quant fits at an 8K context.
- 8 GB VRAM: no quant fits at an 8K context.
- 12 GB VRAM: the best quant that fits is the main file (7.2 GB file, about 9.9 GB in use).
Which TokForge Bonsai 2 27B quant should you choose?
If memory is tight, the smallest file is the first one in the table at 5.9 GB, with the quality trade-offs that come with fewer bits.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about TokForge Bonsai 2 27B
How much VRAM does TokForge Bonsai 2 27B need?
About 9.0 GB at its main file with an 8K context window. That is the 6.4 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.
Can I run TokForge Bonsai 2 27B on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 8.6 GB, so you would need a larger GPU or CPU offloading.
What is the file size of TokForge Bonsai 2 27B GGUF?
It depends on the quant: file is 5.9 GB, file is 6.4 GB, file is 7.2 GB.
What context length does TokForge Bonsai 2 27B support?
The GGUF header lists a trained context length of 262,144 tokens (256K). The KV cache grows with the context you use: about 2.1 GB at 8K and 8.6 GB at 32K.
Where can I download TokForge Bonsai 2 27B GGUF?
The files are on Hugging Face in the darkmaniac7/TokForge-Bonsai-2-27B-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.
What license does TokForge Bonsai 2 27B use?
Hugging Face lists the license as apache-2.0. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.
Can I run TokForge Bonsai 2 27B on 8 GB VRAM?
Yes, the PTQ1_0 file is 5.95 GB and requires about 8.6 GB of VRAM at 8 k context, which fits on 8 GB GPUs.
Does the model support vision?
No, the vision tower is shipped separately as a Q8_0 mmproj pack.
What license does it use?
The model is released under the apache-2.0 license.