Alibaba Qwen · Released July 2025
Qwen3 Coder 30B-A3B Instruct
What it takes to run Qwen3 Coder 30B-A3B Instruct locally: memory by quantization, the GPUs that handle it, and how fast it runs.
Updated
- Parameters
- 30.5B (3.3B active)
- Architecture
- Mixture of experts
- Context
- 256K tokens
- License
- Apache 2.0 license
- Memory at Q4_K_M
- ~21GB
What do you need to run Qwen3 Coder 30B-A3B Instruct?
Qwen3 Coder 30B-A3B Instruct needs about 21GB of VRAM at Q4_K_M with 8K context (19GB for the weights plus KV cache and overhead). The smallest GPU that runs it comfortably at Q3_K_M is a 20GB card like the RTX 4000 Ada. For a good balance of price and speed we recommend the RTX 3090 (~135 tokens/sec). Because it is a mixture-of-experts model with 3.3B active parameters, it generates text much faster than a dense model of the same size, and it tolerates partial CPU offload better.
See the full hardware guide for Qwen3 Coder 30B-A3B Instruct
What to buy
GPUs for Qwen3 Coder 30B-A3B Instruct
Speeds are estimates at the quantization shown with 8K context. We may earn a commission when you buy through these links. Learn more.
Memory
Qwen3 Coder 30B-A3B Instruct VRAM requirements by quantization
Lower-bit quantizations shrink the model at some cost to quality. Q4_K_M is the usual sweet spot. Totals include the KV cache for the context length shown plus runtime overhead.
| Quantization | Quality | Weights | Total @ 8K ctx | Total @ 32K ctx |
|---|---|---|---|---|
| Q2_K | Rough | 12 GB | 13 GB | 16 GB |
| Q3_K_M | Usable | 15 GB | 17 GB | 19 GB |
| Q4_K_M | RecommendedDefault | 19 GB | 21 GB | 23 GB |
| Q5_K_M | Very good | 22 GB | 24 GB | 26 GB |
| Q6_K | Near-lossless | 25 GB | 28 GB | 30 GB |
| Q8_0 | Practically lossless | 33 GB | 35 GB | 37 GB |
| BF16 | Full precision | 61 GB | 64 GB | 67 GB |
KV cache note: Same attention layout as Qwen3 30B-A3B: 2 x 48 layers x 4 KV heads x 128 dims x 2 bytes = 98,304 bytes (96 KiB) per token, about 0.8GB at 8K and 25.8GB at the full 262K context.
Performance
How fast Qwen3 Coder 30B-A3B Instruct runs
Single-user generation speed on popular GPUs that fit the model entirely in memory, at the best quantization each can hold.
RTX 5090
Q6_K · 32GB
~175tok/s
RTX PRO 6000
Q8_0 · 96GB
~150tok/s
RTX 4090
Q4_K_M · 24GB
~150tok/s
RTX 3090
Q4_K_M · 24GB
~135tok/s
RTX PRO 5000 72GB
Q8_0 · 72GB
~125tok/s
RX 7900 XTX
Q4_K_M · 24GB
~110tok/s
RTX 6000 Ada
Q8_0 · 48GB
~99tok/s
V100 32GB
Q6_K · 32GB
~92tok/s
M5 Ultra 512GB
Q8_0 · 512GB
~80tok/s
M5 Ultra 256GB
Q8_0 · 256GB
~80tok/s
Measured
Qwen3 Coder 30B-A3B Instruct benchmarks
- DGX SparkQ8_0 · llama.cpp61 tok/s llama.cpp repository: benches/dgx-spark/dgx-spark.md
Overview
About Qwen3 Coder 30B-A3B Instruct
Qwen3 Coder 30B-A3B is the compact member of Alibaba's 2025 coding line, built for agentic coding tools such as Qwen Code and Cline, with a native 256K context. It became the default local coding model of late 2025 because its 3.3B active parameters keep it fast on modest GPUs. Qwen3.6 35B-A3B has since overtaken it on agentic coding while adding vision and a thinking mode, so it is now superseded. It needs about 21GB at Q4_K_M with 8K context, fitting a 24GB GPU, and it runs well on 12GB to 16GB cards with experts offloaded to system RAM.
Strengths
- +Tuned for repository-scale agentic coding with reliable tool calls
- +Fast: 3.3B active parameters per token
- +256K native context for large codebases
- +Apache 2.0 license
Weaknesses
- −No thinking mode, so it is weaker on hard algorithmic problems
- −Superseded by Qwen3.6 35B-A3B for local agentic coding
- −Narrow focus: general chat and writing lag general-purpose models
Setup
How to run Qwen3 Coder 30B-A3B Instruct
$ ollama run qwen3-coder:30b$ llama-server -hf unsloth/Qwen3-Coder-30B-A3B-Instruct-GGUF:Q4_K_M -c 8192 -ngl 99Weights: Qwen/Qwen3-Coder-30B-A3B-Instruct · GGUF: unsloth/Qwen3-Coder-30B-A3B-Instruct-GGUF
Alternatives
Models like Qwen3 Coder 30B-A3B Instruct
Alibaba Qwen
Qwen3.6 35B-A3B
35B (3B active) · 256K ctx
Alibaba Qwen
Qwen3.8 27B
27B · 256K ctx
Alibaba Qwen
Qwen3 30B-A3B (2507)
30.5B (3.3B active) · 256K ctx
Z.ai
GLM-4.5-Air
106B (12B active) · 128K ctx
Frequently asked questions
How much VRAM do you need to run Qwen3 Coder 30B-A3B Instruct?
About 21GB at Q4_K_M with 8K tokens of context: 18.6GB of weights, 0.8GB of KV cache, and 1.2GB of runtime overhead. Higher-precision quants and longer contexts need more; see the table above.
What is the cheapest GPU that can run Qwen3 Coder 30B-A3B Instruct?
The smallest card in our database that runs it comfortably is the RTX 4000 Ada (20GB), at Q3_K_M and ~84 tokens/sec.
Is Qwen3 Coder 30B-A3B Instruct free for commercial use?
Qwen3 Coder 30B-A3B Instruct is released under the Apache 2.0 license. That is a permissive license, so commercial use is allowed.
How long a context does Qwen3 Coder 30B-A3B Instruct support?
Up to 262,144 tokens (256K). Long contexts need a lot of extra memory for the KV cache, so most local setups run 8K to 32K.
Sources: Qwen3 Coder 30B-A3B Instruct model card (Hugging Face), Unsloth Qwen3 Coder 30B-A3B GGUF files. Capability score basis: Placed below the Qwen3 32B anchor from its Artificial Analysis Intelligence Index v3 standing as a non-reasoning model (October 2025); its coding-specific results are stronger than the general score suggests.