Ling 3.0 Tiny needs about 9.8 GB of memory at Q4_K_M (a 4.8 GB file) with an 8K context window, so it fits on a 12 GB GPU. This repository has 28 quants, from 0.0 GB to 15.8 GB (F16). Longer context needs more: at 32K tokens the KV cache alone adds about 17.7 GB.
Ling 3.0 Tiny – a lightweight hybrid‑reasoning MoE model
Ling 3.0 Tiny is a 7.9 B total‑parameter model with only 1.3 B active parameters per token. It uses a 3:1 KDA–MLA architecture and a sparse MoE FFN with 128 experts, enabling efficient long‑context reasoning and agentic tasks on local hardware.
Good at
- Fast reasoning with low activation
- Coding and instruction following
- Agentic tool‑calling
Limits
- Requires two llama.cpp PRs for full agentic use
- Model may terminate mid‑think
- No vision support
Best for
- Local edge inference with limited VRAM
- Rapid prototyping of reasoning agents
- Educational coding assistants
Who it’s for. Developers and researchers needing a compact reasoning model for local deployment. The model is suitable for systems with modest GPU or CPU resources, such as Apple Silicon Macs or entry‑level NVIDIA GPUs.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| — | 0.0 GB | 5.0 GB | Fits | Get file → | |
| Q1_0 | 1.3 GB | 6.2 GB | Fits | Get file → | |
| IQ1_S | 1.8 GB | 6.7 GB | Fits | Get file → | |
| IQ1_M | 1.9 GB | 6.9 GB | Fits | Get file → | |
| IQ2_XXS | 2.2 GB | 7.1 GB | Fits | Get file → | |
| IQ2_XS | 2.4 GB | 7.4 GB | Tight | Get file → | |
| IQ2_S | 2.5 GB | 7.4 GB | Tight | Get file → | |
| IQ2_M | 2.7 GB | 7.6 GB | Tight | Get file → | |
| Q2_K | 3.0 GB | 7.9 GB | Tight | Get file → | |
| IQ3_XXS | 3.1 GB | 8.1 GB | Too big | Get file → | |
| Q3_K_S | 3.5 GB | 8.4 GB | Too big | Get file → | |
| IQ3_S | 3.5 GB | 8.4 GB | Too big | Get file → | |
| Q3_K_M | 3.8 GB | 8.8 GB | Too big | Get file → | |
| IQ4_XS | 4.3 GB | 9.2 GB | Too big | Get file → | |
| Q4_0 | 4.5 GB | 9.5 GB | Too big | Get file → | |
| Q4_K_S | 4.6 GB | 9.5 GB | Too big | Get file → | |
| MXFP4_MOE | 4.7 GB | 9.6 GB | Too big | Get file → | |
| Q4_K_M | 4.8 GB | 9.8 GB | Too big | Get file → | |
| Q4_K_XL | 5.3 GB | 10.3 GB | Too big | Get file → | |
| Q5_0 | 5.5 GB | 10.4 GB | Too big | Get file → | |
| Q5_K_S | 5.5 GB | 10.4 GB | Too big | Get file → | |
| Q5_K_M | 5.6 GB | 10.6 GB | Too big | Get file → | |
| Q6_K | 6.5 GB | 11.4 GB | Too big | Get file → | |
| Q6_K_XL | 7.3 GB | 12.2 GB | Too big | Get file → | |
| Q8_0 | 8.4 GB | 13.3 GB | Too big | Get file → | |
| Q8_K_XL | 11.2 GB | 16.1 GB | Too big | Get file → | |
| BF16 | 15.8 GB | 20.7 GB | Too big | Get file → | |
| F16 | 15.8 GB | 20.7 GB | Too big | Get file → |
For 8 GB VRAM, use Q4_K_M (≈9.8 GB memory at 8K context) or Q4_K_S (≈9.5 GB). For 12 GB VRAM, Q5_K_M (≈10.6 GB) or Q6_K (≈11.4 GB) are suitable. Larger VRAM allows Q8_0 (≈13.3 GB) or Q8_K_XL (≈16.1 GB).
What GPU do you need for Ling 3.0 Tiny?
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: the best quant that fits is the main file (0.0 GB file, about 5.0 GB in use).
- 8 GB VRAM: the best quant that fits is Q2_K (3.0 GB file, about 7.9 GB in use).
- 12 GB VRAM: the best quant that fits is Q6_K (6.5 GB file, about 11.4 GB in use).
- 16 GB VRAM: the best quant that fits is Q8_0 (8.4 GB file, about 13.3 GB in use).
- 24 GB VRAM: the best quant that fits is BF16 (15.8 GB file, about 20.7 GB in use).
Which Ling 3.0 Tiny quant should you choose?
Q4_K_M is the usual starting point for Ling 3.0 Tiny: it balances file size and quality well. Here it is a 4.8 GB file.
The Q8_0 file is 8.4 GB, about 1.7 times the size of Q4_K_M. It keeps the weights closer to the original, which only matters if you have the memory to spare.
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
Source
Frequently asked questions about Ling 3.0 Tiny
How much VRAM does Ling 3.0 Tiny need?
About 9.8 GB at Q4_K_M with an 8K context window. That is the 4.8 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 Ling 3.0 Tiny on an 8 GB GPU?
Yes, with Q2_K: the 3.0 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 Ling 3.0 Tiny GGUF?
It depends on the quant: file is 0.0 GB, Q1_0 is 1.3 GB, IQ1_S is 1.8 GB, IQ1_M is 1.9 GB, IQ2_XXS is 2.2 GB, IQ2_XS is 2.4 GB and 22 more.
What context length does Ling 3.0 Tiny support?
The GGUF header lists a trained context length of 131,072 tokens (128K). The KV cache grows with the context you use: about 4.4 GB at 8K and 17.7 GB at 32K.
Where can I download Ling 3.0 Tiny GGUF?
The files are on Hugging Face in the bloomer010/Ling-3.0-tiny-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 Ling 3.0 Tiny 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 Ling 3.0 Tiny on 8 GB VRAM?
Yes, with Q4_K_M or Q4_K_S quantization, which use about 9.5–9.8 GB at 8K context. Lower‑precision quants may reduce VRAM usage further.
Does the model support vision tasks?
No, Ling 3.0 Tiny is a text‑only model and does not include vision capabilities.
What is the maximum context length?
The model was trained with 131,072 tokens and supports up to that context length in inference.