Tess 4 27B ROCmFPX needs about 22.2 GB of memory at Q3_0 (a 19.5 GB file) with an 8K context window, so it fits on a 24 GB GPU. This repository has 8 quants, from 15.1 GB (Q3_0) to 28.7 GB (Q8_0). Longer context needs more: at 32K tokens the KV cache alone adds about 8.7 GB.
Tess 4 27B ROCmFPX – a 27.3B reasoning model for AMD GPUs
Tess 4 27B ROCmFPX is a 27.3B parameter model built on Qwen/Qwen3.6-27B. It is post‑trained on 64K‑token long‑context agentic traces and supports reasoning and tool use. The model is distributed in GGUF format for AMD‑native ROCmFPX quantization.
Good at
- Reasoning with long context
- Tool‑based problem solving
- Agentic multi‑step tasks
Limits
- No vision support on standard llama.cpp
- Requires AMD‑native ROCmFPX fork
- Large VRAM needed for higher quants
Best for
- Engineering code analysis
- Complex multi‑step reasoning
- Tool‑integrated workflows
Who it’s for. Developers and researchers needing a large, reasoning‑capable model on AMD hardware. The model is licensed under apache-2.0 and is suitable for applications that can leverage its agentic capabilities.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q3_0 | 15.1 GB | 17.8 GB | Too big | Get file → | |
| Q4_0 | 15.7 GB | 18.4 GB | Too big | Get file → | |
| Q4_0 | 17.7 GB | 20.4 GB | Too big | Get file → | |
| Q3_0 | 19.5 GB | 22.2 GB | Too big | Get file → | |
| Q6_0 | 22.5 GB | 25.2 GB | Too big | Get file → | |
| Q6_0 | 25.3 GB | 28.0 GB | Too big | Get file → | |
| Q8_0 | 28.2 GB | 30.9 GB | Too big | Get file → | |
| Q8_0 | 28.7 GB | 31.3 GB | Too big | Get file → |
For 8 GB VRAM, use Q3_0 (15.13 GB file, 17.8 GB memory at 8 k context). For 12 GB VRAM, Q4_0 (15.72 GB file, 18.4 GB memory). For 16 GB VRAM, Q6_0 (22.53 GB file, 25.2 GB memory). These estimates are approximate.
What GPU do you need for Tess 4 27B ROCmFPX?
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: no quant fits at an 8K context.
- 16 GB VRAM: no quant fits at an 8K context.
- 24 GB VRAM: the best quant that fits is Q3_0 (19.5 GB file, about 22.2 GB in use).
- 32 GB VRAM: the best quant that fits is Q8_0 (28.7 GB file, about 31.3 GB in use).
Which Tess 4 27B ROCmFPX quant should you choose?
Q3_0 (19.5 GB) is the middle option in this repository. A reasonable choice when a 4-bit file will not fit, mostly on larger models. Quality loss is noticeable but often acceptable.
If memory is tight, the smallest file is Q3_0 at 15.1 GB, with the quality trade-offs that come with fewer bits. A reasonable choice when a 4-bit file will not fit, mostly on larger models. Quality loss is noticeable but often acceptable.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about Tess 4 27B ROCmFPX
How much VRAM does Tess 4 27B ROCmFPX need?
About 22.2 GB at Q3_0 with an 8K context window. That is the 19.5 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 Tess 4 27B ROCmFPX on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 17.8 GB, so you would need a larger GPU or CPU offloading.
What is the file size of Tess 4 27B ROCmFPX GGUF?
It depends on the quant: Q3_0 is 15.1 GB, Q4_0 is 15.7 GB, Q4_0 is 17.7 GB, Q3_0 is 19.5 GB, Q6_0 is 22.5 GB, Q6_0 is 25.3 GB and 2 more.
What context length does Tess 4 27B ROCmFPX support?
The GGUF header lists a trained context length of 262,144 tokens (256K). The KV cache grows with the context you use: about 2.2 GB at 8K and 8.7 GB at 32K.
Where can I download Tess 4 27B ROCmFPX GGUF?
The files are on Hugging Face in the lmcoleman/Tess-4-27B-ROCmFPX-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 Tess 4 27B ROCmFPX 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 Tess 4 27B ROCmFPX on 8 GB VRAM?
Only the Q3_0 quant (15.13 GB file, 17.8 GB memory at 8 k context) may fit, but VRAM usage will be close to the limit.
Does this model support vision?
The base Tess-4-27B includes a vision tower, but the ROCmFPX GGUFs do not load on standard llama.cpp and require the AMD fork.
What license applies to Tess 4 27B ROCmFPX?
The model is distributed under apache-2.0, inheriting the base model license.