Menu
mutinaiPreview: Public preview · figures are illustrative fixture data, attributed to their source · data to May 2025 · read-only
What can I run?
Model · Qwen3 · April 2025 Live · Hugging Face

Qwen3 30B-A3B

A 31 billion-parameterHow many numbers the model learned during training — the usual rough measure of its size. mixture-of-expertsA model split into parts where only a few run for each word, so it answers faster than its size suggests. model using ~3.3 billion parameters per token from Qwen Team (Alibaba Cloud). Dense and MoE models with hybrid thinking modes.

Size
30.5B3.3B active
Memory to run
~18 GB estimatesmallest, 8K ctx
Context
32Ktokens
License
PermissiveApache License 2.0
Updated
18 Sept 2026
Compare

Where it runs

Best-fitting version on each reference system, at an 8K contextHow much text the model can consider at once, counted in tokens — roughly ¾ of a word each..

Your hardware →

Runs well 11

Slowly (CPU or offload) 2

Too large 6

Measured throughput

QuantizationSystemRuntimeContextMeasurementsSource
Qwen3 30B-A3B Q4_K_M (Unsloth AI)RTX 4090 workstation (64 GB DDR5)llama.cppcuda · b53004K
3,400 tok/s · Prompt processing (512)
152 tok/s · Generation (128)
Mutinai illustrative fixtures
Qwen3 30B-A3B Q4_K_M (Unsloth AI)Ryzen AI Max+ 395 mini PC (128 GB)llama.cppvulkan · b53004K
420 tok/s · Prompt processing (512)
51 tok/s · Generation (128)
Mutinai illustrative fixtures

Versions & downloads

Each version is a separate set of weights. Expand one for its downloads: quantizationStoring each of the model’s numbers with fewer bits, so the file is smaller and needs less memory. trades a little quality for a much smaller file, and the memory column adds the working memoryExtra space the model needs while it answers. It grows with the length of the conversation, on top of the file itself. a conversation needs on top.

Variant Qwen3 30B-A3B BasePretrained foundation weights — a starting point for fine-tuning, not for chat. · Qwen Team (Alibaba Cloud) · 1 downloadPermissive
Good for
Publisher
Qwen Team (Alibaba Cloud)
License
Apache License 2.0 · commercial use allowed
Released
29 Apr 2025

Contributions are not open yet, so there is nothing here from members.

QuantizationFormatBitsDownloadMemory @ 8KPublisher
Download BF16nativesafetensors1656.9 GiB~60.4 GiB estimateQwen Team (Alibaba Cloud)Qwen/Qwen3-30B-A3B-Base
Variant Qwen3 30B-A3BTuned to follow instructions and hold a conversation. The usual choice. · Qwen Team (Alibaba Cloud) · 8 downloadsPermissive
Good for
ChatReasoningCodingTool use
Publisher
Qwen Team (Alibaba Cloud)
License
Apache License 2.0 · commercial use allowed
Released
29 Apr 2025

Contributions are not open yet, so there is nothing here from members.

QuantizationFormatBitsDownloadMemory @ 8KPublisher
Download BF16nativesafetensors1656.9 GiB~60.4 GiB estimateQwen Team (Alibaba Cloud)Qwen/Qwen3-30B-A3B
Download Q8_0legacy ggufgguf8.530.2 GiB~32.7 GiB estimateLM Studio Communitylmstudio-community/Qwen3-30B-A3B-GGUF
Download MLX 8-bitmlxmlx8.530.2 GiB~32.7 GiB estimateMLX Communitymlx-community/Qwen3-30B-A3B-8bit
Download Q8_0legacy ggufgguf8.530.3 GiB~32.7 GiB estimateQwen Team (Alibaba Cloud)Qwen/Qwen3-30B-A3B-GGUF
Download FP8fp8safetensors8.128.8 GiB~31.2 GiB estimateQwen Team (Alibaba Cloud)Qwen/Qwen3-30B-A3B-FP8
Download Q4_K_Mk quantgguf4.8917.3 GiB~19.2 GiB estimateQwen Team (Alibaba Cloud)Qwen/Qwen3-30B-A3B-GGUF
Download Q4_K_Mk quantgguf4.8917.4 GiB~19.3 GiB estimateUnsloth AIunsloth/Qwen3-30B-A3B-GGUF
Download MLX 4-bitmlxmlx4.516.0 GiB~17.9 GiB estimateMLX Communitymlx-community/Qwen3-30B-A3B-4bit
Variant Qwen3 30B A3B Instruct 2507Tuned to follow instructions and hold a conversation. The usual choice. · Qwen Team (Alibaba Cloud) · 0 downloadsPermissive
Good for
Chat
Publisher
Qwen Team (Alibaba Cloud)
License
Apache License 2.0 · commercial use allowed
Released
28 Jul 2025

Contributions are not open yet, so there is nothing here from members.

No downloadable artifacts recorded.

Variant CantoneseLLM v2.0 30B A3B ThinkingA community or third-party fine-tune of another variant. · hon9kon9ize (third-party) · 0 downloadsPermissive
Good for
Chat
Publisher
hon9kon9ize
License
Apache License 2.0 · commercial use allowed
Released
1 Aug 2026

Contributions are not open yet, so there is nothing here from members.

No downloadable artifacts recorded.

Variant Qwen3 30B A3B ShapleyMCG K34 Validation ReconstructionA community or third-party fine-tune of another variant. · brandonmusic (third-party) · 0 downloadsPermissive
Good for
Chat
Publisher
brandonmusic
License
Apache License 2.0 · commercial use allowed
Released
24 Aug 2026

Contributions are not open yet, so there is nothing here from members.

No downloadable artifacts recorded.

Other sizes in Qwen3: Qwen3 0.6B, Qwen3 1.7B, Qwen3 4B, Qwen3 8B, Qwen3 14B, Qwen3 32B, Qwen3 235B-A22B

Benchmarks

Each benchmarkA fixed set of questions every model is given, so their scores can be compared on the same task. score is labelled with who produced it. Developer-reported scores use each lab’s own prompts and settings; a benchmark that runs every model itself uses one setup for all. Bars are relative to the best open result.

BenchmarkQwen3 30B-A3B
Benchmark GPQA DiamondAccuracy
65.8src
Benchmark LiveCodeBenchpass@1
62.6src

developer-reported

Community

Benchmark runs and reviews from people running it on their own hardware.

Benchmark runs

Contributions are not open yet, so there is nothing here from members.

Reviews

Contributions are not open yet, so there is nothing here from members.

Details & sources

Architecture

Parameters
30.53B
Active / token
3.3B
Architecture
Mixture of experts
Layers
48
Attention heads
32
KV heads
4
Head dim
128
KV cache @ 8K (fp16)
0.75 GiB
Max context
32,768 tokens

Lineage

How this model’s versions relate to each other and to other models.

VariantQwen3 30B-A3B Base base
VariantCantoneseLLM v2.0 30B A3B Thinking fine-tuned from Qwen3 30B-A3B Base
VariantQwen3 30B A3B ShapleyMCG K34 Validation Reconstruction fine-tuned from Qwen3 30B-A3B Base
VariantQwen3 30B-A3B instruct

Sources

  • Hugging Face Hub (live source, 1 record, 13 Sept 2026)
  • Mutinai illustrative fixtures (illustrative fixture, 1 record, 13 Sept 2026)

External identifiers

Field history

  • license = "apache-2.0" — Hugging Face Hub (current)
  • auto_promotion = {"repo":"Qwen/Qwen3-30B-A3B-Instruct-2507","rule":"first_party_release","evidence":["publisher Qwen is Qwen Team (Alibaba Cloud), developer of Qwen, Qwen Coder, Qwen Math","repo name Qwen3-30B-A3B-Instruct-2507 belongs to family Qwen","config.json: 48 layers, 32 attention heads, 4 KV heads, head dim 128, context 262144","safetensors metadata: 30,532,122,624 parameters, consistent with 30B","128 experts (8 per token); A3B active parameters as named by the publisher","published 2025-07-28 (Hugging Face repository creation date)","variant kind instruct stated by the name"]} — Hugging Face Hub (current)
  • base = "Qwen/Qwen3-30B-A3B-Base" — Hugging Face Hub (current)
  • derivation = "fine_tune" — Hugging Face Hub (current)
  • license = "apache-2.0" — Hugging Face Hub (current)
  • name = "CantoneseLLM v2.0 30B A3B Thinking" — Hugging Face Hub (current)
  • base = "Qwen/Qwen3-30B-A3B-Base" — Hugging Face Hub (current)
  • derivation = "fine_tune" — Hugging Face Hub (current)
  • license = "apache-2.0" — Hugging Face Hub (current)
  • name = "Qwen3 30B A3B ShapleyMCG K34 Validation Reconstruction" — Hugging Face Hub (current)

Timeline

  1. Official Qwen3 GGUF quantizations published
  2. Qwen3 released with MoE and hybrid thinking