flint MXFP4 MOE 40B A2.5B needs about 29.6 GB of memory at Q5_K_M (a 28.3 GB file) with an 8K context window, so it fits on a 32 GB GPU. This repository has 3 quants, from 22.2 GB (MXFP4_MOE) to 79.6 GB (BF16). Longer context needs more: at 32K tokens the KV cache alone adds about 3.3 GB.
Flint MXFP4 MOE 40B A2.5B – a 39.8B parameter coding agent model
Flint MXFP4 MOE 40B A2.5B is a 39.8 billion‑parameter mixture‑of‑experts model with 2.5 billion active parameters per token. It supports a 262,144‑token context window, reasoning, and multi‑turn tool use. The model is licensed under Apache‑2.0.
Good at
- Coding tasks and bug fixing
- Tool‑based code generation
- Large‑context reasoning
Limits
- Requires 262,144‑token context, no vision support
- Large VRAM needed for full model
- Apache‑2.0 license restricts commercial redistribution
Best for
- Developing coding assistants
- Automated test generation
- Codebase exploration
Who it’s for. Engineers and researchers building coding agents who need a large‑context, mixture‑of‑experts model. The model is suitable for environments that can support high VRAM and the Apache‑2.0 license terms.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| MXFP4_MOE | 22.2 GB | 23.6 GB | Too big | Get file → | |
| Q5_K_M | 28.3 GB | 29.6 GB | Too big | Get file → | |
| BF16 | 79.6 GB | 80.9 GB | Too big | Get file → |
For 8 GB VRAM use the MXFP4_MOE quant (22.24 GB file, ~23.6 GB memory at 8 k context). For 12 GB VRAM use Q5_K_M (28.31 GB file, ~29.6 GB memory). The BF16 variant (79.63 GB file, ~80.9 GB memory) requires a high‑end GPU.
What GPU do you need for flint MXFP4 MOE 40B A2.5B?
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 MXFP4_MOE (22.2 GB file, about 23.6 GB in use).
- 32 GB VRAM: the best quant that fits is Q5_K_M (28.3 GB file, about 29.6 GB in use).
- 48 GB VRAM: the best quant that fits is Q5_K_M (28.3 GB file, about 29.6 GB in use).
Which flint MXFP4 MOE 40B A2.5B quant should you choose?
Q5_K_M (28.3 GB) is the middle option in this repository. Closer to the original model than 4-bit, for a moderate increase in size. A good pick when you have memory to spare.
If memory is tight, the smallest file is MXFP4_MOE at 22.2 GB, with the quality trade-offs that come with fewer bits. Lower bit counts mean smaller files and more quality loss.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about flint MXFP4 MOE 40B A2.5B
How much VRAM does flint MXFP4 MOE 40B A2.5B need?
About 29.6 GB at Q5_K_M with an 8K context window. That is the 28.3 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 flint MXFP4 MOE 40B A2.5B on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 23.6 GB, so you would need a larger GPU or CPU offloading.
What is the file size of flint MXFP4 MOE 40B A2.5B GGUF?
It depends on the quant: MXFP4_MOE is 22.2 GB, Q5_K_M is 28.3 GB, BF16 is 79.6 GB.
What context length does flint MXFP4 MOE 40B A2.5B support?
The GGUF header lists a trained context length of 262,144 tokens (256K). The KV cache grows with the context you use: about 0.8 GB at 8K and 3.3 GB at 32K.
Where can I download flint MXFP4 MOE 40B A2.5B GGUF?
The files are on Hugging Face in the prithivMLmods/flint-MXFP4_MOE-40B-A2.5B repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.
What license does flint MXFP4 MOE 40B A2.5B 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 Flint MXFP4 MOE on 8 GB VRAM?
Yes, the MXFP4_MOE quant file is 22.24 GB and requires about 23.6 GB of VRAM at 8 k context.
Does Flint support vision tasks?
No, Flint is a decoder‑only language model and does not support vision inputs.
What license applies to this model?
The model is released under the Apache‑2.0 license.