Skip to content
localmodel.run

Text model · Hunyuan

HY Hunyuan-A13B-Instruct: RAM and VRAM requirements

Hunyuan-A13B-Instruct needs about 48.3 GB to run at Q4_K_M (the Q4_K_M GGUF file is ~45.43 GB to download; KV cache and overhead add the rest), or about 82.5 GB at Q8_0. The lightest hardware that runs it is Apple M4 Max (128GB).

Hunyuan family · 80B params (Mixture-of-Experts: activates only 13B of 80B params per token, so generation is faster than the total size suggests) · released Jun 2025.

Shopping for hardware? See what runs Hunyuan-A13B-Instruct →

License tencent-hunyuan-community · Conditional ↓ 50.5K/mo ♥ 796 on HuggingFace
Q4_K_M GGUF
45.43 GB
Q8_0 GGUF
79.58 GB
Memory @ Q4 (4k)
~48.3 GB
Context
256 k

Q4_K_M and Q8_0 are GGUF download (file) sizes. Memory @ Q4 includes the KV cache and overhead, so it runs larger than the file on disk.

Will it run on your device?

Hunyuan-A13B-Instruct runs on 4 of 40 tracked devices at Q4_K_M.

4 run well 0 tight fit 36 too small
Fit check Q4_K_M
No, not enough memory
needs 48.3 GB usable 10.5 GB
See the full breakdown

Memory breakdown

Weights (Q4_K_M)45.43 GB
+
KV cache (4k)2.1 GB
+
Overhead0.8 GB
=
Total48.3 GB

How context length changes it

4k context ~48.3 GB 32k context ~63.4 GB 128k context ~114.9 GB

Longer context grows the KV cache, which for Hunyuan-A13B-Instruct is sized by its 13B active params, not the full 80B. It needs ~48.3 GB at 4k and ~114.9 GB at 128k.

Quantization sizes

GGUF quantson disk
Quantization Size on disk
Q2_K 33.5 GB est
Q3_K_M 39.1 GB est
Q4_K_M (default) 45.43 GB
Q5_K_M 57 GB est
Q6_K 65.6 GB est
Q8_0 79.58 GB
FP16 161 GB

Lower quant = smaller and faster, slightly lower quality. Q4_K_M is the common default.

Run it

llama.cpp
$ llama-cli -hf tencent/Hunyuan-A13B-Instruct-GGUF:Q4_K_M
LM Studio
$ lms get tencent/Hunyuan-A13B-Instruct-GGUF

Which devices can run Hunyuan-A13B-Instruct?

Apple Silicon Macs

RAM-only laptops

iPhone & iPad

Android

NVIDIA GPUs

AMD GPUs

Same job, different size

Pick by what fits your memory: step down to free up VRAM, or step up if you have headroom.

Next size down Qwen2.5 72B 72B Next size up Sarvam-105B 105B

Similar models

FAQ

How much VRAM or RAM does Hunyuan-A13B-Instruct need?

At Q4_K_M, Hunyuan-A13B-Instruct needs about 48.3 GB (weights ~45.43 GB + KV cache + overhead) at a 4k context. At Q8_0 budget ~82.5 GB.

What is the Q4_K_M GGUF file size of Hunyuan-A13B-Instruct?

The Q4_K_M GGUF file is about 45.43 GB to download, and the Q8_0 GGUF is about 79.58 GB. That is the weights file on disk; to run it you also need room for the KV cache and overhead, so budget ~48.3 GB of memory at Q4_K_M.

Can Hunyuan-A13B-Instruct run on a laptop?

Hunyuan-A13B-Instruct is large; you need a high-memory Mac or multi-GPU setup at Q4_K_M.

Is Hunyuan-A13B-Instruct cheaper to run because it is a MoE model?

It is faster, not lighter. Hunyuan-A13B-Instruct activates only 13B of 80B params per token (so it runs quickly), but all experts must stay in memory, so it still needs memory for the full 80B.

Can I use Hunyuan-A13B-Instruct commercially?

Conditionally. Tencent Hunyuan Community License; restricted in the EU and above 100M MAU.

Understand the numbers

Short guides to the ideas behind Hunyuan-A13B-Instruct's memory and quant figures.

Tencent's Hunyuan MoE: 80B total / 13B active, 256K context. Runs on a 64GB+ machine at Q4.

Sources

Last validated 2026-08-03. Memory figures are estimates. See methodology.