Model → hardware
Qwen3-Coder 30B-A3B Instruct
Plan for 24 GB of dedicated GPU memory at Q4_K_M and a 4,000-token context. About 2.5 GB remains after the estimate and reserve.
The full 30.5B weights must fit. Its 3.3B active parameters do not make this a small-memory model.
Coding mixture-of-experts model. Memory fit is an estimate, not a speed or quality guarantee.
Hardware starting points
Planning estimatesQ4_K_M · 32 GB system RAM · one active request · full GPU residency
| Dedicated GPU memory | At 4,000 tokens | At 16,000 tokens | Example from these guides |
|---|---|---|---|
| 24 GB | Smallest listed tier that fits20.8 GB + 0.72 GB reserve | Fits the estimate22.2 GB + reserve | RTX 4090 24 GB |
| 32 GB | More headroom20.8 GB + 0.96 GB reserve | Fits the estimate22.2 GB + reserve | RTX 5090 32 GB |
The smallest tier fits the calculation; it is not a verified minimum. Cache precision and runtime buffers can change the fit. “More headroom” means at least 2 GB remains after the estimate and reserve.
The download is a GGUF conversion; Q5 in the table is an estimate, not a listed file in that repository. Use the publisher’s Instruct / chat template in a current GGUF-capable runtime. The 24 GB tier can become tight when display use, buffers or context grow. CPU offloading needs adequate system RAM and changes performance; it is not included in these full-GPU estimates.
Compare quantisation and context
| File profile | Weight input¹ | GPU estimate at 4K | GPU estimate at 16K |
|---|---|---|---|
| Q4_K_M | 19.50 GB | 20.8 GB | 22.2 GB |
| Q5_K_M | 23.80 GB | 25.1 GB | 26.5 GB |
¹ Catalogue planning weights. These can differ from an exact downloaded file; runtime memory also includes cache and buffers.
Keep the exact file, runtime version, cache type and context limit consistent when comparing. Q5 uses more memory; compare its answers on your own tasks before accepting the trade-off.
Setup and common questions
Why does A3B need so much memory?
Qwen lists 30.5B total parameters and 3.3B activated. Routing selects experts during inference, but full GPU residency still needs the model’s full weights. Active parameters are not the weight footprint and do not establish measured speed.
Which runtime and file should I use?
LM Studio Community provides a matching Q4_K_M GGUF conversion, about 18.60 GB before cache and buffers. Use a current runtime with this MoE architecture and tool-call format. The original checkpoint is Apache 2.0 and non-thinking-only. The Q5 table is a planning profile; the linked repository has no Q5 file.
Can I load an entire repository?
The publisher lists 262,144 native tokens. Our 4K and 16K estimates do not verify that window. Increase context only after checking available memory, prompt-processing time and whether the model handles your repository’s task correctly.
How are these recommendations chosen?
We compare dedicated GPU memory tiers using the main Finder’s memory calculation at the stated quantisation and context. These tiers describe memory fit, not speed or output quality.
GPU memory includes weights, context cache and a 0.9 GB runtime allowance, plus a 3% reserve (0.15–1.5 GB). Cache format and runtime buffers can change the fit. See the calculation assumptions.
Sources and related guides
Specification sources checked 3 October 2026.
These comparisons are planning estimates. No reviewed or independently reproduced measurements are available yet. About test evidence.