Skip to content
localmodel.run

Text model · Qwen3-Coder

Qwen3-Coder-Next 80B-A3B: RAM and VRAM requirements

Qwen3-Coder-Next 80B-A3B needs about 48 GB to run at Q4_K_M (the Q4_K_M GGUF file is ~45.09 GB to download; KV cache and overhead add the rest), or about 81.9 GB at Q8_0. The lightest hardware that runs it is Apple M4 Max (64GB).

Qwen3-Coder family · 80B params (Mixture-of-Experts: activates only 3B of 80B params per token, so generation is faster than the total size suggests) · released 2026-02-02.

Shopping for hardware? See what runs Qwen3-Coder-Next 80B-A3B →

License Apache-2.0 · Commercial OK ↓ 596.4K/mo ♥ 1.7K on HuggingFace
Q4_K_M GGUF
45.09 GB
Q8_0 GGUF
78.99 GB
Memory @ Q4 (4k)
~48 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?

Qwen3-Coder-Next 80B-A3B runs on 5 of 43 tracked devices at Q4_K_M.

4 run well 1 tight fit 38 too small
Fit check Q4_K_M
No, not enough memory
needs 48 GB usable 10.5 GB
See the full breakdown

Memory breakdown

Weights (Q4_K_M)45.09 GB
+
KV cache (4k)2.1 GB
+
Overhead0.8 GB
=
Total48 GB

How context length changes it

4k context ~48 GB 32k context ~63.1 GB 128k context ~114.6 GB

Longer context grows the KV cache, which for Qwen3-Coder-Next 80B-A3B is sized by its 3B active params, not the full 80B. It needs ~48 GB at 4k and ~114.6 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) 45.09 GB
Q5_K_M 57 GB est
Q6_K 65.6 GB est
Q8_0 78.99 GB
FP16 159.46 GB

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

Run it

Ollama
$ ollama run qwen3-coder-next:latest
llama.cpp
$ llama-cli -hf Qwen/Qwen3-Coder-Next-GGUF:Q4_K_M
LM Studio
$ lms get Qwen/Qwen3-Coder-Next-GGUF

Which devices can run Qwen3-Coder-Next 80B-A3B?

NVIDIA GPUs

Apple Silicon Macs

RAM-only laptops

iPhone & iPad

Android

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

Head-to-head

FAQ

How much VRAM or RAM does Qwen3-Coder-Next 80B-A3B need?

At Q4_K_M, Qwen3-Coder-Next 80B-A3B needs about 48 GB (weights ~45.09 GB + KV cache + overhead) at a 4k context. At Q8_0 budget ~81.9 GB.

What is the Q4_K_M GGUF file size of Qwen3-Coder-Next 80B-A3B?

The Q4_K_M GGUF file is about 45.09 GB to download, and the Q8_0 GGUF is about 78.99 GB. That is the weights file on disk; to run it you also need room for the KV cache and overhead, so budget ~48 GB of memory at Q4_K_M.

Can Qwen3-Coder-Next 80B-A3B run on a laptop?

Qwen3-Coder-Next 80B-A3B is large; you need a high-memory Mac or multi-GPU setup at Q4_K_M.

Is Qwen3-Coder-Next 80B-A3B cheaper to run because it is a MoE model?

It is faster, not lighter. Qwen3-Coder-Next 80B-A3B activates only 3B 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 Qwen3-Coder-Next 80B-A3B commercially?

Yes. Qwen3-Coder-Next 80B-A3B is licensed Apache-2.0, which permits commercial use.

Understand the numbers

Short guides to the ideas behind Qwen3-Coder-Next 80B-A3B's memory and quant figures.

Coding MoE with 80B total and 3B active parameters, 256K native context, and Apache-2.0. Q4_K_M, Q8_0 and F16 sizes are exact four-shard sums from Qwen's official GGUF repository. Ollama latest has a separately packed 51.74 GB model layer.

Sources

Catalog updated 2026-10-05. Memory figures are estimates. See methodology.