GP

gpt oss 6.0b specialized all pruned moe only 7 experts i1

gpt-ossGGUF by mradermacherUpdated 2025-12-04
Download on Hugging Face
Parameters5.98B
Context128K
Layers24
Architecturegpt-oss
Smallest quant0.0 GB
Quick answer

gpt oss 6.0b specialized all pruned moe only 7 experts i1 needs about 5.5 GB of memory at Q4_K_M (a 4.6 GB file) with an 8K context window, so it fits on a 6 GB GPU. This repository has 25 quants, from 0.0 GB to 6.3 GB (Q6_K). Longer context needs more: at 32K tokens the KV cache alone adds about 1.6 GB.

Pruned GPT-OSS 6.0B with 7 experts per layer

This is a pruned variant of GPT-OSS-20B, reduced to 7 experts per layer. The model retains broad capabilities while cutting the expert parameter count to 21.9% of the original. It is experimental and may not perform consistently across all tasks.

Good at

  • Reasoning
  • Tools usage
  • General-purpose language tasks

Limits

  • Experimental model
  • May not perform well on all benchmarks
  • No vision support

Best for

  • Edge deployment
  • Resource‑constrained inference
  • Research on MoE pruning

Who it’s for. Developers and researchers looking for a smaller GPT‑OSS base model for experimentation or deployment on devices with limited memory. The model is suitable for those who can tolerate variability in performance due to pruning.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
—0.0 GB 0.9 GB
Fits Get file →
IQ1_M3.6 GB 4.5 GB
Fits Get file →
IQ1_S3.6 GB 4.5 GB
Fits Get file →
IQ2_XXS3.6 GB 4.5 GB
Fits Get file →
IQ2_XS3.6 GB 4.5 GB
Fits Get file →
Q2_K_S3.6 GB 4.5 GB
Fits Get file →
Q3_K_S3.6 GB 4.5 GB
Fits Get file →
IQ2_M3.6 GB 4.5 GB
Fits Get file →
IQ2_S3.6 GB 4.5 GB
Fits Get file →
IQ3_S3.6 GB 4.5 GB
Fits Get file →
IQ3_XS3.6 GB 4.5 GB
Fits Get file →
IQ3_XXS3.6 GB 4.5 GB
Fits Get file →
Q2_K3.6 GB 4.5 GB
Fits Get file →
IQ4_XS3.7 GB 4.6 GB
Fits Get file →
IQ4_NL3.7 GB 4.6 GB
Fits Get file →
Q4_03.7 GB 4.6 GB
Fits Get file →
IQ3_M3.7 GB 4.6 GB
Fits Get file →
Q3_K_M3.9 GB 4.8 GB
Fits Get file →
Q3_K_L4.0 GB 4.9 GB
Fits Get file →
Q4_14.0 GB 4.9 GB
Fits Get file →
Q4_K_S4.3 GB 5.2 GB
Fits Get file →
Q4_K_M4.6 GB 5.5 GB
Fits Get file →
Q5_K_S4.7 GB 5.6 GB
Fits Get file →
Q5_K_M4.9 GB 5.8 GB
Fits Get file →
Q6_K6.3 GB 7.2 GB
Tight Get file →

For 8 GB VRAM, the Q6_K quant (file size 6.3 GB, estimated memory 7.2 GB at 8 k context) fits comfortably. If you need a smaller file, Q4_K_M (4.58 GB, 5.5 GB memory) or Q3_K_L (3.98 GB, 4.9 GB memory) are good alternatives. Use the IQ‑quants for potentially better quality at similar sizes.

What GPU do you need for gpt oss 6.0b specialized all pruned moe only 7 experts i1?

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 (0.0 GB file, about 0.9 GB in use).
  • 6 GB VRAM: the best quant that fits is Q5_K_M (4.9 GB file, about 5.8 GB in use).
  • 8 GB VRAM: the best quant that fits is Q6_K (6.3 GB file, about 7.2 GB in use).

Which gpt oss 6.0b specialized all pruned moe only 7 experts i1 quant should you choose?

Q4_K_M is the usual starting point for gpt oss 6.0b specialized all pruned moe only 7 experts i1: it balances file size and quality well. Here it is a 4.6 GB file.

If memory is tight, the smallest file is the first one in the table at 0.0 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

Architecturegpt-oss
Parameters5.98B (approximate, mixture of experts)
Layers24
Trained context131,072 tokens
KV cache (fp16)48 KB per token

Frequently asked questions about gpt oss 6.0b specialized all pruned moe only 7 experts i1

How much VRAM does gpt oss 6.0b specialized all pruned moe only 7 experts i1 need?

About 5.5 GB at Q4_K_M with an 8K context window. That is the 4.6 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 gpt oss 6.0b specialized all pruned moe only 7 experts i1 on an 8 GB GPU?

Yes, with Q6_K: the 6.3 GB file needs about 7.2 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of gpt oss 6.0b specialized all pruned moe only 7 experts i1 GGUF?

It depends on the quant: file is 0.0 GB, IQ1_M is 3.6 GB, IQ1_S is 3.6 GB, IQ2_XXS is 3.6 GB, IQ2_XS is 3.6 GB, Q2_K_S is 3.6 GB and 19 more.

What context length does gpt oss 6.0b specialized all pruned moe only 7 experts i1 support?

The GGUF header lists a trained context length of 131,072 tokens (128K). The KV cache grows with the context you use: about 0.4 GB at 8K and 1.6 GB at 32K.

Where can I download gpt oss 6.0b specialized all pruned moe only 7 experts i1 GGUF?

The files are on Hugging Face in the mradermacher/gpt-oss-6.0b-specialized-all-pruned-moe-only-7-experts-i1-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 gpt oss 6.0b specialized all pruned moe only 7 experts i1 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 this model on 8 GB VRAM?

Yes, the Q6_K quant requires an estimated 7.2 GB at 8 k context, which fits within 8 GB VRAM.

Does this model support vision tasks?

No, the model does not include vision capabilities.

Is the model fully fine‑tuned?

No, it is a pruned variant without additional training.

Other GGUFs of gpt-oss-6.0b-specialized-all-pruned-moe-only-7-experts

Similar models

Browse all →

More from mradermacher

All by mradermacher →

More gpt-oss models

All gpt-oss →

Other small models

All small models →