SP

SPEED 1 BASE

SpeedGGUF by mradermacherUpdated 2026-10-17
Download on Hugging Face
Parameters19.2B
Context4K
Layers48
Architecturellama
Smallest quant0.0 GB
Quick answer

SPEED 1 BASE needs about 13.7 GB of memory at Q4_K_M (a 11.6 GB file) with an 8K context window, so it fits on a 16 GB GPU. This repository has 24 quants, from 0.0 GB to 15.7 GB (Q6_K).

SPEED 1 BASE – 19B parameter MoE language model

SPEED 1 BASE is a 19.2B parameter mixture‑of‑experts model built by SKT AI LABS. It is trained for complex reasoning, chain‑of‑thought tasks, mathematics, and coding. The model is available in multiple GGUF quantizations.

Good at

  • Complex reasoning
  • Chain‑of‑thought tasks
  • Mathematics and coding

Limits

  • Requires 8k context
  • Large VRAM for higher quants
  • No vision support

Best for

  • Developers needing reasoning models
  • Researchers testing MoE architectures
  • Applications needing chain‑of‑thought generation

Who it’s for. Engineers and researchers who need a large, reasoning‑capable model and have the GPU memory to run it. The model is open‑source under the MIT license, making it suitable for both academic and commercial use.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
—0.0 GB 1.3 GB
Fits Get file →
IQ1_S4.1 GB 5.4 GB
Fits Get file →
IQ1_M4.5 GB 5.8 GB
Fits Get file →
IQ2_XXS5.1 GB 6.4 GB
Fits Get file →
IQ2_XS5.7 GB 7.0 GB
Fits Get file →
IQ2_S5.9 GB 7.2 GB
Fits Get file →
IQ2_M6.4 GB 7.7 GB
Tight Get file →
Q2_K_S6.6 GB 7.9 GB
Tight Get file →
Q2_K7.1 GB 8.4 GB
Too big Get file →
IQ3_XXS7.4 GB 8.7 GB
Too big Get file →
IQ3_XS7.9 GB 9.2 GB
Too big Get file →
Q3_K_S8.3 GB 9.6 GB
Too big Get file →
IQ3_S8.3 GB 9.6 GB
Too big Get file →
IQ3_M8.5 GB 9.8 GB
Too big Get file →
Q3_K_M9.2 GB 10.5 GB
Too big Get file →
Q3_K_L10.0 GB 11.3 GB
Too big Get file →
IQ4_XS10.3 GB 11.6 GB
Too big Get file →
Q4_010.9 GB 12.2 GB
Too big Get file →
Q4_K_S10.9 GB 12.2 GB
Too big Get file →
Q4_K_M11.6 GB 12.9 GB
Too big Get file →
Q4_112.0 GB 13.3 GB
Too big Get file →
Q5_K_S13.2 GB 14.5 GB
Too big Get file →
Q5_K_M13.6 GB 14.9 GB
Too big Get file →
Q6_K15.7 GB 17.0 GB
Too big Get file →

For 8‑GB VRAM, use IQ1_S (file size 4.05 GB, memory 6.2 GB). For 12‑GB VRAM, IQ3_S (8.33 GB, 10.4 GB). For 16‑GB VRAM, Q4_K_M (11.58 GB, 13.7 GB). Estimates are approximate.

What GPU do you need for SPEED 1 BASE?

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 2.1 GB in use).
  • 6 GB VRAM: the best quant that fits is the main file (0.0 GB file, about 2.1 GB in use).
  • 8 GB VRAM: the best quant that fits is IQ2_S (5.9 GB file, about 8.0 GB in use).
  • 12 GB VRAM: the best quant that fits is Q3_K_M (9.2 GB file, about 11.3 GB in use).
  • 16 GB VRAM: the best quant that fits is Q5_K_M (13.6 GB file, about 15.7 GB in use).
  • 24 GB VRAM: the best quant that fits is Q6_K (15.7 GB file, about 17.9 GB in use).

Which SPEED 1 BASE quant should you choose?

Q4_K_M is the usual starting point for SPEED 1 BASE: it balances file size and quality well. Here it is a 11.6 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

Architecturellama
Parameters19.2B (approximate, mixture of experts)
Layers48
Trained context4,096 tokens
KV cache (fp16)192 KB per token

Source

Quantized bymradermacher
FamilySpeed
Licensemit
Downloads (HF)1,011
Last updated2026-10-17

Frequently asked questions about SPEED 1 BASE

How much VRAM does SPEED 1 BASE need?

About 13.7 GB at Q4_K_M with an 8K context window. That is the 11.6 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 SPEED 1 BASE on an 8 GB GPU?

Yes, with IQ2_S: the 5.9 GB file needs about 8.0 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of SPEED 1 BASE GGUF?

It depends on the quant: file is 0.0 GB, IQ1_S is 4.1 GB, IQ1_M is 4.5 GB, IQ2_XXS is 5.1 GB, IQ2_XS is 5.7 GB, IQ2_S is 5.9 GB and 18 more.

What context length does SPEED 1 BASE support?

The GGUF header lists a trained context length of 4,096 tokens (4K). The KV cache grows with the context you use: about 1.6 GB at 8K and 0.8 GB at 4K.

Where can I download SPEED 1 BASE GGUF?

The files are on Hugging Face in the mradermacher/SPEED-1-BASE-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 SPEED 1 BASE use?

Hugging Face lists the license as mit. 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 SPEED 1 BASE on 8 GB VRAM?

Use the IQ1_S quant, which requires about 6.2 GB of VRAM at 8k context.

Does SPEED 1 BASE support vision tasks?

No, the model is a text‑only language model.

Other GGUFs of SPEED-1-BASE

Similar models

Browse all →

More from mradermacher

All by mradermacher →

More Speed models

All Speed →

Other medium models

All medium models →