RI

Ring Lite

RingGGUF by mradermacherUpdated 2025-06-21
Download on Hugging Face
Parameters16.8B
Context32K
Layers28
Architecturebailingmoe
Smallest quant5.4 GB
Quick answer

Ring Lite needs about 12.1 GB of memory at Q4_K_M (a 11.2 GB file) with an 8K context window, so it fits on a 16 GB GPU. This repository has 24 quants, from 5.4 GB (IQ1_S) to 15.1 GB (Q6_K). Longer context needs more: at 32K tokens the KV cache alone adds about 1.9 GB.

Ring Lite: 16.8B parameter MoE LLM with 32K context

Ring Lite is a lightweight Mixture of Experts LLM based on Ring-lite. It has 16.8B parameters and supports a 32768 token context window. The model is released under the MIT license and is fully open‑source.

Good at

  • Complex reasoning
  • Math and code tasks
  • Instruction following

Limits

  • Requires large VRAM for higher quants
  • No vision support
  • Limited to text only

Best for

  • Educational math tutoring
  • Code generation assistance
  • Research prototyping

Who it’s for. Developers and researchers who need a reasoning‑capable LLM with a large context window but want a smaller memory footprint. The model is suitable for local deployment on GPUs with sufficient VRAM.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
IQ1_S5.4 GB 6.3 GB
Fits Get file →
IQ1_M5.6 GB 6.6 GB
Fits Get file →
IQ2_XXS6.1 GB 7.0 GB
Fits Get file →
IQ2_XS6.4 GB 7.4 GB
Tight Get file →
IQ2_S6.5 GB 7.4 GB
Tight Get file →
IQ2_M6.8 GB 7.8 GB
Tight Get file →
Q2_K6.9 GB 7.9 GB
Tight Get file →
Q2_K_S7.0 GB 7.9 GB
Tight Get file →
IQ3_XXS7.5 GB 8.4 GB
Too big Get file →
IQ3_XS7.6 GB 8.6 GB
Too big Get file →
Q3_K_S8.0 GB 9.0 GB
Too big Get file →
IQ3_S8.0 GB 9.0 GB
Too big Get file →
IQ3_M8.1 GB 9.1 GB
Too big Get file →
Q3_K_M8.7 GB 9.7 GB
Too big Get file →
Q3_K_L9.1 GB 10.0 GB
Too big Get file →
IQ4_XS9.2 GB 10.2 GB
Too big Get file →
IQ4_NL9.5 GB 10.5 GB
Too big Get file →
Q4_09.6 GB 10.5 GB
Too big Get file →
Q4_K_S10.2 GB 11.2 GB
Too big Get file →
Q4_110.6 GB 11.5 GB
Too big Get file →
Q4_K_M11.2 GB 12.1 GB
Too big Get file →
Q5_K_S11.9 GB 12.9 GB
Too big Get file →
Q5_K_M12.7 GB 13.7 GB
Too big Get file →
Q6_K15.1 GB 16.0 GB
Too big Get file →

For 8 GB VRAM, use IQ1_S (5.38 GB, 6.3 GB memory at 8K context) or IQ1_M (5.64 GB, 6.6 GB). For 12 GB VRAM, IQ2_M (6.8 GB, 7.8 GB) or Q2_K (6.91 GB, 7.9 GB) are suitable. For 16 GB VRAM, IQ3_M (8.12 GB, 9.1 GB) or Q3_K_M (8.73 GB, 9.7 GB) provide a balance of size and performance.

What GPU do you need for Ring Lite?

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 Q2_K_S (7.0 GB file, about 7.9 GB in use).
  • 12 GB VRAM: the best quant that fits is Q4_1 (10.6 GB file, about 11.5 GB in use).
  • 16 GB VRAM: the best quant that fits is Q6_K (15.1 GB file, about 16.0 GB in use).

Which Ring Lite quant should you choose?

Q4_K_M is the usual starting point for Ring Lite: it balances file size and quality well. Here it is a 11.2 GB file.

If memory is tight, the smallest file is IQ1_S at 5.4 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

Architecturebailingmoe
Parameters16.8B (approximate, mixture of experts)
Layers28
Trained context32,768 tokens
KV cache (fp16)56 KB per token

Source

Quantized bymradermacher
FamilyRing
Licensemit
Downloads (HF)560
Last updated2025-06-21

Frequently asked questions about Ring Lite

How much VRAM does Ring Lite need?

About 12.1 GB at Q4_K_M with an 8K context window. That is the 11.2 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 Ring Lite on an 8 GB GPU?

Yes, with Q2_K_S: the 7.0 GB file needs about 7.9 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of Ring Lite GGUF?

It depends on the quant: IQ1_S is 5.4 GB, IQ1_M is 5.6 GB, IQ2_XXS is 6.1 GB, IQ2_XS is 6.4 GB, IQ2_S is 6.5 GB, IQ2_M is 6.8 GB and 18 more.

What context length does Ring Lite 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 Ring Lite GGUF?

The files are on Hugging Face in the mradermacher/Ring-lite-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 Ring Lite 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 Ring Lite on 8 GB VRAM?

Yes, using the IQ1_S or IQ1_M quantized files, which are 5.38 GB and 5.64 GB respectively and require about 6.3–6.6 GB VRAM at 8K context.

Does Ring Lite support vision tasks?

No, Ring Lite is a text‑only model and does not include vision capabilities.

Other GGUFs of Ring-lite

Similar models

Browse all →

More from mradermacher

All by mradermacher →

More Ring models

All Ring →

Other medium models

All medium models →