QW

Qwen3.6 35B A3B TinyKV

QwenGGUF by darioooooo0oUpdated 2026-10-26
Download on Hugging Face
Parameters35.5B
Context256K
Layers41
Architectureqwen35moe
Smallest quant12.6 GB
Quick answer

Qwen3.6 35B A3B TinyKV needs about 23.2 GB of memory at Q4_K_M (a 22.1 GB file) with an 8K context window, so it fits on a 24 GB GPU. This repository has 8 quants, from 12.6 GB (IQ2_M) to 37.8 GB (Q8_0). Longer context needs more: at 32K tokens the KV cache alone adds about 2.8 GB.

Qwen3.6 35B A3B TinyKV – 35.5B parameter LLM with reduced KV cache

Qwen3.6 35B A3B TinyKV is a 35.5B parameter model with a 2.5× smaller KV cache. It retains the original MoE experts and vision tower, but only the 10 full‑attention layers were retrained. The model is available in multiple GGUF quantizations.

Good at

  • Reasoning
  • Tool usage
  • Large context handling

Limits

  • Requires ≥80 GB GPU for bf16
  • Only text inference tested
  • No vision inputs tested

Best for

  • Concurrent 32k‑token request workloads
  • Memory‑constrained inference
  • LLM‑powered tooling

Who it’s for. Developers needing a large‑parameter model with efficient KV usage on a single GPU. Suitable for applications that require reasoning and tool integration but can tolerate the 80 GB GPU requirement.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
IQ2_M12.6 GB 13.8 GB
Too big Get file →
Q3_K_M17.6 GB 18.8 GB
Too big Get file →
IQ4_XS19.5 GB 20.7 GB
Too big Get file →
MXFP4_MOE20.3 GB 21.5 GB
Too big Get file →
Q4_K_M22.1 GB 23.2 GB
Too big Get file →
Q5_K_M25.6 GB 26.8 GB
Too big Get file →
Q6_K29.4 GB 30.6 GB
Too big Get file →
Q8_037.8 GB 39.0 GB
Too big Get file →

For 8 GB VRAM, use IQ2_M (12.6 GB file, ~13.8 GB memory). For 16 GB VRAM, Q3_K_M (17.6 GB file, ~18.8 GB memory). For 32 GB VRAM, Q4_K_M (22.06 GB file, ~23.2 GB memory).

What GPU do you need for Qwen3.6 35B A3B TinyKV?

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: no quant fits at an 8K context.
  • 12 GB VRAM: no quant fits at an 8K context.
  • 16 GB VRAM: the best quant that fits is IQ2_M (12.6 GB file, about 13.8 GB in use).
  • 24 GB VRAM: the best quant that fits is Q4_K_M (22.1 GB file, about 23.2 GB in use).
  • 32 GB VRAM: the best quant that fits is Q6_K (29.4 GB file, about 30.6 GB in use).
  • 48 GB VRAM: the best quant that fits is Q8_0 (37.8 GB file, about 39.0 GB in use).

Which Qwen3.6 35B A3B TinyKV quant should you choose?

Q4_K_M is the usual starting point for Qwen3.6 35B A3B TinyKV: it balances file size and quality well. Here it is a 22.1 GB file.

The Q8_0 file is 37.8 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 IQ2_M at 12.6 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

Architectureqwen35moe
Parameters35.5B (approximate, mixture of experts)
Active parameters3B
Layers41
Trained context262,144 tokens
KV cache (fp16)82 KB per token

Source

Quantized bydarioooooo0o
FamilyQwen
Licenseapache-2.0
Downloads (HF)1,303
Last updated2026-10-26

Frequently asked questions about Qwen3.6 35B A3B TinyKV

How much VRAM does Qwen3.6 35B A3B TinyKV need?

About 23.2 GB at Q4_K_M with an 8K context window. That is the 22.1 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 Qwen3.6 35B A3B TinyKV on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 13.8 GB, so you would need a larger GPU or CPU offloading.

What is the file size of Qwen3.6 35B A3B TinyKV GGUF?

It depends on the quant: IQ2_M is 12.6 GB, Q3_K_M is 17.6 GB, IQ4_XS is 19.5 GB, MXFP4_MOE is 20.3 GB, Q4_K_M is 22.1 GB, Q5_K_M is 25.6 GB and 2 more.

What context length does Qwen3.6 35B A3B TinyKV support?

The GGUF header lists a trained context length of 262,144 tokens (256K). The KV cache grows with the context you use: about 0.7 GB at 8K and 2.8 GB at 32K.

Where can I download Qwen3.6 35B A3B TinyKV GGUF?

The files are on Hugging Face in the darioooooo0o/Qwen3.6-35B-A3B-TinyKV-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 Qwen3.6 35B A3B TinyKV 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 Qwen3.6 35B A3B TinyKV on 8 GB VRAM?

Use the IQ2_M quant, which requires about 13.8 GB memory at 8k context. It will fit on 8 GB VRAM with some swapping.

Does the model support vision inputs?

The vision tower is unchanged, but image/video inputs were not tested with the new attention. Use with caution.

What GPU is needed for bf16 weights?

A single GPU with at least 80 GB VRAM is required for bf16 inference.

Similar models

Browse all →

More from darioooooo0o

All by darioooooo0o →

More Qwen models

All Qwen →

Other large models

All large models →