FL

Flint 40B A2.5B Abliterated

FlintGGUF by mradermacherUpdated 2026-10-27
Download on Hugging Face
Parameters39.8B
Context256K
Layers49
Architectureflint
Smallest quant14.7 GB
Quick answer

Flint 40B A2.5B Abliterated needs about 25.5 GB of memory at Q4_K_M (a 24.2 GB file) with an 8K context window, so it fits on a 32 GB GPU. This repository has 11 quants, from 14.7 GB (Q2_K) to 42.4 GB (Q8_0). Longer context needs more: at 32K tokens the KV cache alone adds about 3.3 GB.

Flint 40B A2.5B Abliterated is a 39.8B-parameter Flint model in GGUF format, quantized by mradermacher. Its trained context length is 256K tokens. The repository has 11 quant files, from 14.7 GB (Q2_K) to 42.4 GB (Q8_0).

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q2_K14.7 GB 16.0 GB
Too big Get file →
Q3_K_S17.3 GB 18.6 GB
Too big Get file →
Q3_K_M19.1 GB 20.4 GB
Too big Get file →
Q3_K_L20.7 GB 22.0 GB
Too big Get file →
IQ4_XS21.5 GB 22.8 GB
Too big Get file →
Q4_K_S22.7 GB 24.0 GB
Too big Get file →
Q4_K_M24.2 GB 25.5 GB
Too big Get file →
Q5_K_S27.5 GB 28.8 GB
Too big Get file →
Q5_K_M28.3 GB 29.6 GB
Too big Get file →
Q6_K32.7 GB 34.0 GB
Too big Get file →
Q8_042.4 GB 43.7 GB
Too big Get file →

What GPU do you need for Flint 40B A2.5B Abliterated?

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: the best quant that fits is Q2_K (14.7 GB file, about 16.0 GB in use).
  • 24 GB VRAM: the best quant that fits is Q4_K_S (22.7 GB file, about 24.0 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 Q8_0 (42.4 GB file, about 43.7 GB in use).

Which Flint 40B A2.5B Abliterated quant should you choose?

Q4_K_M is the usual starting point for Flint 40B A2.5B Abliterated: it balances file size and quality well. Here it is a 24.2 GB file.

The Q8_0 file is 42.4 GB, about 1.8 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 Q2_K at 14.7 GB, with the quality trade-offs that come with fewer bits. Use it only when memory is very tight. Expect a clear drop in quality, which is usually worse on small models and easier to live with on large ones.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architectureflint
Parameters39.8B (approximate, mixture of experts)
Active parameters2.5B
Layers49
Trained context262,144 tokens
KV cache (fp16)98 KB per token

Source

Quantized bymradermacher
FamilyFlint
Licenseapache-2.0
Downloads (HF)553
Last updated2026-10-27

Frequently asked questions about Flint 40B A2.5B Abliterated

How much VRAM does Flint 40B A2.5B Abliterated need?

About 25.5 GB at Q4_K_M with an 8K context window. That is the 24.2 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 40B A2.5B Abliterated on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 16.0 GB, so you would need a larger GPU or CPU offloading.

What is the file size of Flint 40B A2.5B Abliterated GGUF?

It depends on the quant: Q2_K is 14.7 GB, Q3_K_S is 17.3 GB, Q3_K_M is 19.1 GB, Q3_K_L is 20.7 GB, IQ4_XS is 21.5 GB, Q4_K_S is 22.7 GB and 5 more.

What context length does Flint 40B A2.5B Abliterated 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 40B A2.5B Abliterated GGUF?

The files are on Hugging Face in the mradermacher/Flint-40B-A2.5B-abliterated-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 Flint 40B A2.5B Abliterated 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.

Other GGUFs of Flint-40B-A2.5B-abliterated

Similar models

Browse all →

More from mradermacher

All by mradermacher →

More Flint models

All Flint →

Other large models

All large models →