llmrun
AlibabaQwen 3Qwen3MoeForCausalLM

Qwen3 30B A3B Thinking 2507 — Hardware Requirements & GPU Compatibility

ChatFunctions

Qwen3 30B A3B Thinking 2507 is the reasoning-focused variant of Alibaba's 30-billion-parameter mixture-of-experts model, updated in July 2025. Like its instruct sibling, it activates only about 3 billion parameters per token, keeping resource demands low while enabling multi-step reasoning and chain-of-thought problem solving. This thinking variant is designed for tasks that benefit from deliberate, step-by-step logic such as math, coding puzzles, and analytical questions. Its efficient MoE design means users with modest GPUs can still access strong reasoning capabilities without needing datacenter-class hardware.

30.5B params · MoE3.4B active262K contextApache 2.070.8K downloads391 likes37.1K quant downloads
S

Runs great on a 24 GB GPU

Runs from12 GB IQ2_M · 16KBest on24 GB Q4_K_M · ~173 tok/s · 64KMacs from16 GB IQ2_M · ~56 tok/s

Score 100.7#55 of 106 open#9 of 36 that fit 24 GBApache 2.0

Reference GPU · RTX 4090 24 GB

Estimates, ±20% · how we estimate

SOn your Tesla M40 24GB: Q4_K_M fits · 4.6 GB free · ~73 tok/s · up to 64K contextTesla M40 24GB page →

llmrun Composite Score →

100.7

Benchmarks · tier chart ↓
Among 106 open modelsHow the Composite Score is built

Run Qwen3 30B A3B Thinking 2507

Best fit for your Tesla M40 24GB. Every other method, OS and quant is in the step-by-step guide.Each way opens a step-by-step guide for your OS and quant.

Recommended for your Tesla M40 24GB

Ollama
one command · needs Ollama installed
$ollama run qwen3:30b

Q4_K_M fits fully · 4.6 GB free · 16K context

Download: Qwen3-30B-A3B-Thinking-2507-Q4_K_M.gguf18.56 GBunsloth

install, OS, other quants

Ways to runOther ways

each opens step-by-step

Run in cloud

Runs on RTX 3090 (24 GB) (4 GB headroom) · Q4_K_M

Generation speed
~166 tok/s
generation speed
Cost per 1M output tokens
$0.33
per 1M output tokens
Deployon Vast.aiBest valuefrom $0.20/hr (opens in new tab)
Compare GPUs →
or

How Much VRAM Does Qwen3 30B A3B Thinking 2507 Need?

Select a quantization level; the listed hardware adheres to it.

QuantizationVRAM+32KOn your device
13.4 GB14.9 GB2-bit · K-quant — 2-bit quantization with K-quant improvementsFits: 128K context · ~95 tok/s
13.8 GB15.3 GB3-bit small — 3-bit small quantizationFits: 128K context · ~93 tok/s
15.3 GB16.8 GB3-bit medium — 3-bit medium quantizationFits: 128K context · ~86 tok/s
15.7 GB17.2 GB4-bit · legacy — 4-bit legacy quantizationFits: 128K context · ~85 tok/s
18.7 GB20.2 GB4-bit medium — 4-bit medium quantization — most popular sweet spotFits: 64K context · ~73 tok/s
22.1 GB23.7 GB5-bit medium — 5-bit medium quantization — good quality/size tradeoffFits: 32K context · ~64 tok/s
25.6 GB27.1 GB6-bit, very good — 6-bit quantization, very good qualityDoes not fit: needs 26.3 GB
30.9 GB32.4 GB8-bit, near-lossless — 8-bit quantization, near-losslessDoes not fit: needs 31.6 GB

Selected Q4_K_M

4-bit medium quantization

19.4GBat 16K context

Weights
18.3 GB
KV cache
0.8 GB
Overhead
0.3 GB

Fits Tesla M40 24GB · 4.6 GB free · ~73 tok/s

≥50 tok/s needs ≥182 GB/s · RTX 5090 ~227, RTX 3090 Ti ~173

Which Hardware Can Run Qwen3 30B A3B Thinking 2507?

Q4_K_M · 18.7 GB

Qwen3 30B A3B Thinking 2507 (Q4_K_M) requires 18.7 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 25+ GB is recommended. Using the full 262K context window can add up to 12.8 GB, bringing total usage to 31.5 GB. 8 GPUs can run it, including NVIDIA GeForce RTX 5090, NVIDIA GeForce RTX 3090 Ti. 122 devices with unified memory can run Qwen3 30B A3B Thinking 2507, including Mac Studio (M3 Ultra, 256GB), Mac Studio (M3 Ultra, 512GB), Mac Mini M5 Pro (24 GB).

By VRAM tier · typical GPUs and Macs at Q4_K_M
8 GB · RTX 4060Doesn't fit — smallest quant needs 9.5 GB8 GB tier →
12 GB · RTX 3060 12GBIQ2_M · 11.4 GB · ~130 tok/s · 16K12 GB tier →
16 GB · RTX 5060 Ti 16GBQ3_K_M · 16 GB · ~119 tok/s · 16K16 GB tier →
24 GB · RTX 4090referenceQ4_K_M · 19.4 GB · ~173 tok/s · 64K24 GB tier →
Mac 32 GB · Mac Mini M4Q4_K_M · 19.4 GB · ~29 tok/s · 128KMacs →
Mac 64 GB · Mac Studio M4 MaxQ4_K_M · 19.4 GB · ~67 tok/s · 256KMacs →

How good is Qwen3 30B A3B Thinking 2507?

The llmrun Score merges 7 public benchmark boards into a single figure, where 100 represents the average for open models.

Score vs VRAM in the 24 GB tier

  • Qwen3 30B A3B Thinking 2507
  • Other models
  • Best Score for its VRAM

#9 of 36 that fit 24 GBEvery model here fits your Tesla M40 24GB (24 GB)

Each dot is a model at the best quant that fits a 24 GB GPU. The line steps up to the best Score available at each amount of VRAM; a dot below it is beaten by a model that needs no more memory.

Full 24 GB ranking →

Benchmarks

The 6 boards that weigh most in the llmrun Score, of 7 with a result for this model; badges show who ran the test.

  • GPQA Diamond

    Epoch AI · Jul 2025Independent

    70.1%

    #35/100 open

    Independent
  • AIME 2024/2025

    Epoch AI · Jul 2025Independent

    70.3%

    #30/88 open

    Independent
  • MMLU-Pro

    TIGER-LabIndependent

    80.9%

    #22/146 open

    Independent
  • SciCode

    Epoch AI · Jul 2025Lab-reported

    33.0%

    #56/72 open

    Lab-reported
  • CritPt

    Epoch AI · Jul 2025Lab-reported

    0.3%

    #49/84 open

    Lab-reported
  • DTBench

    Epoch AI · Jul 2025Lab-reported

    69.3%

    #35/70 open

    Lab-reported
All 7 Qwen3 30B A3B Thinking 2507 benchmarks

In the 24 GB class

36 models fit 24 GB

Same Score, only models that fit 24 GB at their best quant: the ranking a 24 GB owner actually chooses from.

  1. #1

    Qwen3.8 27BAlibaba

    122.0

    ~38 tok/s · 17.9 GB

  2. #2

    Qwen3.6 27BAlibaba

    111.2

    ~38 tok/s · 17.9 GB

  3. #3

    Qwen3.6 35B A3BAlibaba

    106.8

    ~151 tok/s · 22.0 GB

  4. #4

    Gemma 4 31B ITGoogle

    106.7

    ~32 tok/s · 21.7 GB

  5. #5

    Gemma 4 26B A4B ITGoogle

    105.8

    ~171 tok/s · 16.5 GB

  6. #6

    Qwen3.5 35B A3BAlibaba

    105.8

    ~151 tok/s · 22.0 GB

Full 24 GB ranking →RTX 4090 page →

Alternatives to Qwen3 30B A3B Thinking 2507

Qwen 3 family · 79 models →

Official models with an llmrun Score, each answering one question; numbers are at the best quant that fits a 24 GB GPU.

Similar models

10
ModelScoreVRAMtok/sWhy it is similar
Qwen3.6 27BAlibaba111.217.9 GB~38similar score
Qwen3.6 35B A3BAlibaba106.822 GB~151similar score
Gemma 4 26B A4B ITGoogle105.816.5 GB~171similar score
Qwen3.5 35B A3BAlibaba105.822 GB~151similar score
Qwen3.5 27BAlibaba103.817.9 GB~38similar score
Seed OSS 36B InstructByteDance-Seed102.522.8 GB~31similar score
Qwen3.5 9BAlibaba99.46.6 GB~103similar score
Qwen3.5 4BAlibaba98.53.4 GB~201similar score
DeepSeek R1 Distill Qwen 32BDeepSeek97.223 GB~34same size
QwQ 32BAlibaba96.223 GB~34same size
Fine-tunes & variants(2)

Specifications

Publisher
Alibaba
Family
Qwen 3
Parameters
30.5B
Active parameters
3.4B per token
Architecture
Qwen3MoeForCausalLM
Context Length
262,144 tokens
Vocabulary Size
151,936
Release Date
2025-07-29
License
Apache 2.0

Resources

Frequently Asked Questions

How much VRAM does Qwen3 30B A3B Thinking 2507 need?

Qwen3 30B A3B Thinking 2507 requires 18.7 GB of VRAM at Q4_K_M, or 61.5 GB at BF16. Full 262K context adds up to 12.8 GB (31.5 GB total).

VRAM = Weights + KV Cache + Overhead

Weights = 30.5B × 4.8 bits ÷ 8 = 18.3 GB

KV Cache + Overhead ≈ 0.4 GB (at 2K context + ~0.3 GB framework)

Fit ratings and hardware model lists check this model with room for a 16K-token context, which needs a little more memory.

KV Cache + Overhead ≈ 13.2 GB (at full 262K context)

VRAM usage by quantization

18.7 GB
31.5 GB

Learn more about VRAM estimation →

Can NVIDIA GeForce RTX 4090 run Qwen3 30B A3B Thinking 2507?

Yes, at Q5_K_L (22.5 GB) or lower. Higher quantizations like Q6_K (25.6 GB) exceed the NVIDIA GeForce RTX 4090's 24 GB.

What's the best quantization for Qwen3 30B A3B Thinking 2507?

For Qwen3 30B A3B Thinking 2507, Q4_K_M (18.7 GB) offers the best balance of quality and VRAM usage. Q4_K_L (19.1 GB) provides better quality if you have the VRAM. The smallest option is IQ2_XXS at 8.8 GB.

VRAM requirement by quantization

IQ2_XXS
8.8 GB
IQ3_S
13.4 GB
Q3_K_L
16.1 GB
Q4_K_M ★
18.7 GB
Q5_0
19.5 GB
BF16
61.5 GB

★ Recommended — best balance of quality and VRAM usage.

Learn more about quantization →

Can I run Qwen3 30B A3B Thinking 2507 on a Mac?

Yes — Mac Mini M2 Pro (16 GB) and 133 other Macs can run Qwen3 30B A3B Thinking 2507, the smallest at IQ2_M. Apple Silicon uses unified memory, so the model shares RAM with the system. At Q4_K_M you need at least 18.7 GB of usable unified memory (RAM minus macOS overhead), from Mac Mini M5 Pro (24 GB).

Can I run Qwen3 30B A3B Thinking 2507 locally?

Yes — Qwen3 30B A3B Thinking 2507 can run locally on consumer hardware. Popular tools include Ollama, LM Studio, and llama.cpp.

How fast is Qwen3 30B A3B Thinking 2507?

At Q4_K_M, Qwen3 30B A3B Thinking 2507 can reach ~100 tok/s on AMD Instinct MI350X. On NVIDIA GeForce RTX 4090: ~173 tok/s. Speed depends mainly on GPU memory bandwidth. Real-world results typically within ±20%.

tok/s = 1000 ÷ (active GB ÷ (bandwidth GB/s × efficiency) × 1000 + layers × routing ms)

Mixture-of-Experts: just the active experts get read for each token, along with a fixed routing cost per layer.

Example: NVIDIA B200 → 2.1 GB active ÷ (8000 × 0.65) = 0.39535 ms, plus 48 layers × 0.055 ms = 2.64000 ms, so 1000 ÷ 3.03535 ms = ~329 tok/s

Estimated speed at Q4_K_M (18.7 GB)

~330 tok/s
~173 tok/s
~330 tok/s
~304 tok/s

Real-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.

Learn more about tok/s estimation →

What's the download size of Qwen3 30B A3B Thinking 2507?

At Q4_K_M, the download is about 18.32 GB. The full-precision BF16 version is 61.06 GB. The smallest option (IQ2_XXS) is 8.40 GB.

Which GPUs can run Qwen3 30B A3B Thinking 2507?

8 consumer GPUs can run Qwen3 30B A3B Thinking 2507 at Q4_K_M. Top options include NVIDIA GeForce RTX 5090, AMD Radeon RX 7900 XT, AMD Radeon RX 7900 XTX. 1 GPU has plenty of headroom for comfortable inference.

Which devices can run Qwen3 30B A3B Thinking 2507?

122 devices with unified memory can run Qwen3 30B A3B Thinking 2507 at Q4_K_M (18.7 GB), including AMD Ryzen AI 9 HX 370 (Strix Point) Laptop, ASUS Ascent GX10, Beelink GTR9 Pro (Ryzen AI Max+ 395, 128 GB), Framework Desktop (Ryzen AI Max+ 395, 128 GB). Apple Silicon Macs use unified memory shared between CPU and GPU, making them well-suited for local LLM inference.