SU

Supra2 100M Instruct

SupraGGUF by SupraLabsUpdated 2026-08-05
Download on Hugging Face
Parameters101M
Context2K
Layers12
Architectureqwen3
Smallest quant0.2 GB
Quick answer

Supra2 100M Instruct needs about 0.8 GB of memory at F16 (a 0.2 GB file) with an 8K context window, so it fits on a 4 GB GPU.

Supra2 100M Instruct – a small, instruction‑tuned Qwen3 model

Supra2 100M Instruct is a 100 million‑parameter decoder‑only language model built on the Qwen3 architecture. It was fine‑tuned for instruction following on a single RTX 5060 Ti 16 GB. The model supports up to 2 048 context tokens and is released under Apache‑2.0.

Good at

  • Instruction following on short prompts
  • Fast iteration on limited hardware
  • Educational research on small‑scale pretraining

Limits

  • Only English text
  • No safety or refusal tuning
  • Short context – 2 048 tokens max

Best for

  • Rapid prototyping of instruction models
  • Academic studies of scaling behavior
  • Fine‑tuning on domain‑specific data

Who it’s for. Researchers and developers needing a lightweight, instruction‑tuned model for experimentation. It is not suitable for production or user‑facing deployments.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
F160.2 GB 0.7 GB
Fits Get file →

The F16 quantization file is 0.2 GB and requires about 0.8 GB of VRAM at 8 k context. This is suitable for GPUs with 4 GB or more VRAM. No other quantizations are available.

What GPU do you need for Supra2 100M Instruct?

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 F16 (0.2 GB file, about 0.8 GB in use).

Which Supra2 100M Instruct quant should you choose?

F16 (0.2 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.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architectureqwen3
Parameters101M
Layers12
Trained context2,048 tokens
KV cache (fp16)12 KB per token

Source

Quantized bySupraLabs
FamilySupra
Licenseapache-2.0
Downloads (HF)102,237
Last updated2026-08-05

Frequently asked questions about Supra2 100M Instruct

How much VRAM does Supra2 100M Instruct need?

About 0.8 GB at F16 with an 8K context window. That is the 0.2 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 Supra2 100M Instruct on an 8 GB GPU?

Yes, with F16: the 0.2 GB file needs about 0.8 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of Supra2 100M Instruct GGUF?

It depends on the quant: F16 is 0.2 GB.

What context length does Supra2 100M Instruct support?

The GGUF header lists a trained context length of 2,048 tokens (2K). The KV cache grows with the context you use: about 0.1 GB at 8K and 0.0 GB at 2K.

Where can I download Supra2 100M Instruct GGUF?

The files are on Hugging Face in the SupraLabs/Supra2-100M-Instruct repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.

What license does Supra2 100M Instruct 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 Supra2 100M Instruct on 4 GB VRAM?

Yes, the F16 quantized model needs roughly 0.8 GB of VRAM at 8 k context, well within a 4 GB GPU.

Does this model support vision or multimodal input?

No, it is a purely text‑based decoder‑only model with no vision capabilities.

Is the model safe for user‑facing applications?

No, it lacks safety or refusal tuning and can continue harmful prompts.

Other GGUFs of Supra2-100M-Base

Similar models

Browse all →

More from SupraLabs

All by SupraLabs →

More Supra models

All Supra →

Other tiny models

All tiny models →