IN

amd.Instella MoE 16B A3B Think

InstellaGGUF by DevQuasarUpdated 2026-07-31
Download on Hugging Face
Parameters15.9B
Context32K
Layers27
Architectureinstella-moe
Smallest quant6.5 GB
Quick answer

amd.Instella MoE 16B A3B Think needs about 11.4 GB of memory at F16 (a 10.5 GB file) with an 8K context window, so it fits on a 12 GB GPU. This repository has 6 quants, from 6.5 GB (F16) to 16.9 GB (F16). Longer context needs more: at 32K tokens the KV cache alone adds about 1.9 GB.

amd.Instella MoE 16B A3B Think – a 16B Mixture‑of‑Experts language model

amd.Instella MoE 16B A3B Think is a 15.9B parameter model with 3B active parameters per token. It supports up to 32768 context tokens and is trained on AMD Instinct GPUs using the Primus framework.

Good at

  • reasoning
  • long‑context generation
  • Mixture‑of‑Experts efficiency

Limits

  • no vision support
  • requires large VRAM
  • license is other

Best for

  • research prototyping
  • high‑throughput inference
  • long‑context tasks

Who it’s for. Developers and researchers who need a large, open Mixture‑of‑Experts model with extended context. The model is available through the DevQuasar GitHub repository and can be run with llama.cpp.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
F166.5 GB 7.5 GB
Tight Get file →
F168.2 GB 9.2 GB
Too big Get file →
F1610.5 GB 11.4 GB
Too big Get file →
F1612.0 GB 12.9 GB
Too big Get file →
F1614.2 GB 15.2 GB
Too big Get file →
F1616.9 GB 17.8 GB
Too big Get file →

For 8 GB VRAM, use the F16 quant with file size 6.51 GB; it estimates 7.5 GB memory at 8 k context. For 12 GB VRAM, the 10.47 GB F16 file uses about 11.4 GB memory. Larger VRAM allows the 14.2 GB or 16.87 GB F16 files.

What GPU do you need for amd.Instella MoE 16B A3B Think?

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: the best quant that fits is F16 (6.5 GB file, about 7.5 GB in use).
  • 12 GB VRAM: the best quant that fits is F16 (10.5 GB file, about 11.4 GB in use).
  • 16 GB VRAM: the best quant that fits is F16 (14.2 GB file, about 15.2 GB in use).
  • 24 GB VRAM: the best quant that fits is F16 (16.9 GB file, about 17.8 GB in use).

Which amd.Instella MoE 16B A3B Think quant should you choose?

F16 (10.5 GB) is the middle option in this repository. The unquantized model. It gives original quality and needs the most memory, so it is mainly for conversion or reference.

If memory is tight, the smallest file is F16 at 6.5 GB, with the quality trade-offs that come with fewer bits. The unquantized model. It gives original quality and needs the most memory, so it is mainly for conversion or reference.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architectureinstella-moe
Parameters15.9B (approximate, mixture of experts)
Active parameters3B
Layers27
Trained context32,768 tokens
KV cache (fp16)56 KB per token

Source

Quantized byDevQuasar
FamilyInstella
Licenseother
Downloads (HF)1,704
Last updated2026-07-31

Frequently asked questions about amd.Instella MoE 16B A3B Think

How much VRAM does amd.Instella MoE 16B A3B Think need?

About 11.4 GB at F16 with an 8K context window. That is the 10.5 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 amd.Instella MoE 16B A3B Think on an 8 GB GPU?

Yes, with F16: the 6.5 GB file needs about 7.5 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of amd.Instella MoE 16B A3B Think GGUF?

It depends on the quant: F16 is 6.5 GB, F16 is 8.2 GB, F16 is 10.5 GB, F16 is 12.0 GB, F16 is 14.2 GB, F16 is 16.9 GB.

What context length does amd.Instella MoE 16B A3B Think support?

The GGUF header lists a trained context length of 32,768 tokens (32K). The KV cache grows with the context you use: about 0.5 GB at 8K and 1.9 GB at 32K.

Where can I download amd.Instella MoE 16B A3B Think GGUF?

The files are on Hugging Face in the DevQuasar/amd.Instella-MoE-16B-A3B-Think-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 amd.Instella MoE 16B A3B Think use?

Hugging Face lists the license as other. 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?

The smallest F16 file is 6.51 GB and estimates 7.5 GB memory at 8 k context, so 8 GB VRAM is the minimum.

Does this model support vision?

No, the model has no vision capabilities.

What context length does it support?

It supports up to 32768 tokens.

Other GGUFs of Instella-MoE-16B-A3B-Think

Similar models

Browse all →

More from DevQuasar

All by DevQuasar →

More Instella models

All Instella →

Other medium models

All medium models →