Frontis MA1 35B needs about 23.1 GB of memory at Q4_K_M (a 21.9 GB file) with an 8K context window, so it fits on a 24 GB GPU. This repository has 28 quants, from 0.2 GB to 71.1 GB (BF16). Longer context needs more: at 32K tokens the KV cache alone adds about 2.8 GB.
Frontis MA1 35B – a 35.5B parameter multimodal LLM with execution‑grounded training
Frontis MA1 35B is a 35.5B parameter model trained with OpenMLE for evolutionary program synthesis. It supports text and image input and includes MTP layers for speculative decoding. The model is released under a cc‑by‑nc‑4.0 license.
Good at
- Execution‑grounded program synthesis
- Multimodal text and image understanding
- Speculative decoding via MTP
Limits
- No full vision validation
- Large memory requirement for high context
- License restricts commercial use
Best for
- Research on evolutionary ML
- Code generation with debugging support
- Multimodal dialogue systems
Who it’s for. Researchers and developers working on evolutionary machine learning and program synthesis. The model is suitable for academic projects that require multimodal input and speculative decoding.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| — | 0.2 GB | 1.4 GB | Fits | Get file → | |
| IQ2_XXS | 10.3 GB | 11.4 GB | Too big | Get file → | |
| IQ2_XS | 11.3 GB | 12.5 GB | Too big | Get file → | |
| IQ2_S | 11.5 GB | 12.7 GB | Too big | Get file → | |
| IQ2_M | 12.5 GB | 13.7 GB | Too big | Get file → | |
| Q2_K | 13.1 GB | 14.3 GB | Too big | Get file → | |
| Q2_K_L | 13.6 GB | 14.8 GB | Too big | Get file → | |
| IQ3_XXS | 15.3 GB | 16.5 GB | Too big | Get file → | |
| Q3_K_S | 16.0 GB | 17.2 GB | Too big | Get file → | |
| IQ3_XS | 16.7 GB | 17.9 GB | Too big | Get file → | |
| Q3_K_M | 16.7 GB | 17.9 GB | Too big | Get file → | |
| Q3_K_L | 17.4 GB | 18.5 GB | Too big | Get file → | |
| IQ3_M | 17.4 GB | 18.6 GB | Too big | Get file → | |
| Q3_K_XL | 17.8 GB | 19.0 GB | Too big | Get file → | |
| IQ4_XS | 19.3 GB | 20.5 GB | Too big | Get file → | |
| IQ4_NL | 20.3 GB | 21.5 GB | Too big | Get file → | |
| Q4_0 | 20.4 GB | 21.6 GB | Too big | Get file → | |
| Q4_K_S | 21.1 GB | 22.3 GB | Too big | Get file → | |
| Q4_K_M | 21.9 GB | 23.1 GB | Too big | Get file → | |
| Q4_K_L | 22.2 GB | 23.4 GB | Too big | Get file → | |
| Q4_1 | 22.4 GB | 23.6 GB | Too big | Get file → | |
| Q5_K_S | 24.6 GB | 25.8 GB | Too big | Get file → | |
| Q5_K_M | 25.5 GB | 26.7 GB | Too big | Get file → | |
| Q5_K_L | 25.8 GB | 27.0 GB | Too big | Get file → | |
| Q6_K | 30.5 GB | 31.7 GB | Too big | Get file → | |
| Q6_K_L | 30.8 GB | 32.0 GB | Too big | Get file → | |
| Q8_0 | 37.8 GB | 39.0 GB | Too big | Get file → | |
| BF16 | 71.1 GB (2 parts) | 72.3 GB | Too big | Get file → |
For 8 GB VRAM, use IQ2_XXS (10.26 GB, 11.4 GB memory at 8k context). For 16 GB VRAM, Q4_K_M (21.86 GB, 23.1 GB memory) offers a good balance. For 24 GB VRAM, Q5_K_M (25.49 GB, 26.7 GB memory) provides higher quality.
What GPU do you need for Frontis MA1 35B?
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: the best quant that fits is the main file (0.2 GB file, about 1.4 GB in use).
- 6 GB VRAM: the best quant that fits is the main file (0.2 GB file, about 1.4 GB in use).
- 8 GB VRAM: the best quant that fits is the main file (0.2 GB file, about 1.4 GB in use).
- 12 GB VRAM: the best quant that fits is IQ2_XXS (10.3 GB file, about 11.4 GB in use).
- 16 GB VRAM: the best quant that fits is Q2_K_L (13.6 GB file, about 14.8 GB in use).
- 24 GB VRAM: the best quant that fits is Q4_1 (22.4 GB file, about 23.6 GB in use).
- 32 GB VRAM: the best quant that fits is Q6_K_L (30.8 GB file, about 32.0 GB in use).
- 48 GB VRAM: the best quant that fits is Q8_0 (37.8 GB file, about 39.0 GB in use).
Which Frontis MA1 35B quant should you choose?
Q4_K_M is the usual starting point for Frontis MA1 35B: it balances file size and quality well. Here it is a 21.9 GB file.
The Q8_0 file is 37.8 GB, about 1.7 times the size of Q4_K_M. It keeps the weights closer to the original, which only matters if you have the memory to spare.
If memory is tight, the smallest file is the first one in the table at 0.2 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 Frontis MA1 35B
How much VRAM does Frontis MA1 35B need?
About 23.1 GB at Q4_K_M with an 8K context window. That is the 21.9 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 Frontis MA1 35B on an 8 GB GPU?
Yes, with the smaller files: the 0.2 GB file needs about 1.4 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of Frontis MA1 35B GGUF?
It depends on the quant: file is 0.2 GB, IQ2_XXS is 10.3 GB, IQ2_XS is 11.3 GB, IQ2_S is 11.5 GB, IQ2_M is 12.5 GB, Q2_K is 13.1 GB and 22 more.
What context length does Frontis MA1 35B support?
The GGUF header lists a trained context length of 262,144 tokens (256K). The KV cache grows with the context you use: about 0.7 GB at 8K and 2.8 GB at 32K.
Where can I download Frontis MA1 35B GGUF?
The files are on Hugging Face in the bartowski/FrontisAI_Frontis-MA1-35B-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 Frontis MA1 35B use?
Hugging Face lists the license as cc-by-nc-4.0. 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 Frontis MA1 35B on 8 GB VRAM?
Use the IQ2_XXS quant which is 10.26 GB and requires about 11.4 GB memory at 8k context.
Does the model support image input?
Yes, it supports image input with the provided mmproj projector files.
Is the model suitable for commercial deployment?
The cc‑by‑nc‑4.0 license prohibits commercial use without permission.