geekradar
Int
Explore GeekRadar

More to explore

Good gear All topics About us Say hello
GeekRadar / The comparison desk
COMPARE.
AI models
alibaba

Qwen3-30B-A3B-Instruct-2507

AI model Released

Context window
262,144 tokens
Maximum output
16,384 tokens
Total parameters
30.53 B
Open weights
Yes

Capability & preference

Human preference · Arena

Overall text preferences from head-to-head votes. Not factual accuracy or a task-success rate.

Qwen3-30B-A3B-Instruct-2507 1383.8 qwen3-30b-a3b-instruct-2507 · 95% interval 1379.0–1388.7 · 23,322 votes · 2026-10-15

Recorded results

Only the same benchmark, version, workload, scenario and unit share a row. Missing results are not zero. System results include the full submitted hardware and software configuration.

Arena text preference / chinesehuman preference · 2026-10-15 · rating

Qwen3-30B-A3B-Instruct-2507

1,432.66 rating

Confidence interval: 1,418.30 – 1,447.02

1755 votes

Category
chinese
Evaluated configuration
qwen3-30b-a3b-instruct-2507
Evaluation type
human preference
Arena text preference / codinghuman preference · 2026-10-15 · rating

Qwen3-30B-A3B-Instruct-2507

1,415.92 rating

Confidence interval: 1,407.12 – 1,424.72

4614 votes

Category
coding
Evaluated configuration
qwen3-30b-a3b-instruct-2507
Evaluation type
human preference
Arena text preference / creative_writinghuman preference · 2026-10-15 · rating

Qwen3-30B-A3B-Instruct-2507

1,317.24 rating

Confidence interval: 1,306.32 – 1,328.16

2964 votes

Category
creative_writing
Evaluated configuration
qwen3-30b-a3b-instruct-2507
Evaluation type
human preference
Arena text preference / englishhuman preference · 2026-10-15 · rating

Qwen3-30B-A3B-Instruct-2507

1,391.15 rating

Confidence interval: 1,384.61 – 1,397.69

9759 votes

Category
english
Evaluated configuration
qwen3-30b-a3b-instruct-2507
Evaluation type
human preference
Arena text preference / exclude_tieshuman preference · 2026-10-15 · rating

Qwen3-30B-A3B-Instruct-2507

1,357.88 rating

Confidence interval: 1,351.19 – 1,364.56

16355 votes

Category
exclude_ties
Evaluated configuration
qwen3-30b-a3b-instruct-2507
Evaluation type
human preference
Arena text preference / experthuman preference · 2026-10-15 · rating

Qwen3-30B-A3B-Instruct-2507

1,396.33 rating

Confidence interval: 1,379.67 – 1,413.00

1173 votes

Category
expert
Evaluated configuration
qwen3-30b-a3b-instruct-2507
Evaluation type
human preference
Arena text preference / frenchhuman preference · 2026-10-15 · rating

Qwen3-30B-A3B-Instruct-2507

1,418.83 rating

Confidence interval: 1,386.83 – 1,450.83

336 votes

Category
french
Evaluated configuration
qwen3-30b-a3b-instruct-2507
Evaluation type
human preference
Arena text preference / germanhuman preference · 2026-10-15 · rating

Qwen3-30B-A3B-Instruct-2507

1,378.77 rating

Confidence interval: 1,353.05 – 1,404.50

523 votes

Category
german
Evaluated configuration
qwen3-30b-a3b-instruct-2507
Evaluation type
human preference
Arena text preference / hard_promptshuman preference · 2026-10-15 · rating

Qwen3-30B-A3B-Instruct-2507

1,397.79 rating

Confidence interval: 1,391.51 – 1,404.06

10978 votes

Category
hard_prompts
Evaluated configuration
qwen3-30b-a3b-instruct-2507
Evaluation type
human preference
Arena text preference / hard_prompts_englishhuman preference · 2026-10-15 · rating

Qwen3-30B-A3B-Instruct-2507

1,406.42 rating

Confidence interval: 1,397.72 – 1,415.13

4749 votes

Category
hard_prompts_english
Evaluated configuration
qwen3-30b-a3b-instruct-2507
Evaluation type
human preference
Arena text preference / industry_business_and_management_and_financial_operationshuman preference · 2026-10-15 · rating

Qwen3-30B-A3B-Instruct-2507

1,389.18 rating

Confidence interval: 1,379.83 – 1,398.54

4119 votes

Category
industry_business_and_management_and_financial_operations
Evaluated configuration
qwen3-30b-a3b-instruct-2507
Evaluation type
human preference
Arena text preference / industry_entertainment_and_sports_and_mediahuman preference · 2026-10-15 · rating

Qwen3-30B-A3B-Instruct-2507

1,334.55 rating

Confidence interval: 1,325.43 – 1,343.68

4127 votes

Category
industry_entertainment_and_sports_and_media
Evaluated configuration
qwen3-30b-a3b-instruct-2507
Evaluation type
human preference
Arena text preference / industry_legal_and_governmenthuman preference · 2026-10-15 · rating

Qwen3-30B-A3B-Instruct-2507

1,386.51 rating

Confidence interval: 1,371.13 – 1,401.89

1463 votes

Category
industry_legal_and_government
Evaluated configuration
qwen3-30b-a3b-instruct-2507
Evaluation type
human preference
Arena text preference / industry_life_and_physical_and_social_sciencehuman preference · 2026-10-15 · rating

Qwen3-30B-A3B-Instruct-2507

1,392.85 rating

Confidence interval: 1,383.07 – 1,402.64

3678 votes

Category
industry_life_and_physical_and_social_science
Evaluated configuration
qwen3-30b-a3b-instruct-2507
Evaluation type
human preference
Arena text preference / industry_mathematicalhuman preference · 2026-10-15 · rating

Qwen3-30B-A3B-Instruct-2507

1,411.46 rating

Confidence interval: 1,395.01 – 1,427.91

1208 votes

Category
industry_mathematical
Evaluated configuration
qwen3-30b-a3b-instruct-2507
Evaluation type
human preference
Arena text preference / industry_medicine_and_healthcarehuman preference · 2026-10-15 · rating

Qwen3-30B-A3B-Instruct-2507

1,392.34 rating

Confidence interval: 1,375.98 – 1,408.70

1327 votes

Category
industry_medicine_and_healthcare
Evaluated configuration
qwen3-30b-a3b-instruct-2507
Evaluation type
human preference
Arena text preference / industry_software_and_it_serviceshuman preference · 2026-10-15 · rating

Qwen3-30B-A3B-Instruct-2507

1,416.62 rating

Confidence interval: 1,409.53 – 1,423.70

7962 votes

Category
industry_software_and_it_services
Evaluated configuration
qwen3-30b-a3b-instruct-2507
Evaluation type
human preference
Arena text preference / industry_writing_and_literature_and_languagehuman preference · 2026-10-15 · rating

Qwen3-30B-A3B-Instruct-2507

1,335.95 rating

Confidence interval: 1,327.49 – 1,344.40

5074 votes

Category
industry_writing_and_literature_and_language
Evaluated configuration
qwen3-30b-a3b-instruct-2507
Evaluation type
human preference
Arena text preference / instruction_followinghuman preference · 2026-10-15 · rating

Qwen3-30B-A3B-Instruct-2507

1,362.99 rating

Confidence interval: 1,355.16 – 1,370.82

5971 votes

Category
instruction_following
Evaluated configuration
qwen3-30b-a3b-instruct-2507
Evaluation type
human preference
Arena text preference / japanesehuman preference · 2026-10-15 · rating

Qwen3-30B-A3B-Instruct-2507

1,337.71 rating

Confidence interval: 1,315.63 – 1,359.78

767 votes

Category
japanese
Evaluated configuration
qwen3-30b-a3b-instruct-2507
Evaluation type
human preference
Arena text preference / koreanhuman preference · 2026-10-15 · rating

Qwen3-30B-A3B-Instruct-2507

1,330.70 rating

Confidence interval: 1,305.86 – 1,355.53

593 votes

Category
korean
Evaluated configuration
qwen3-30b-a3b-instruct-2507
Evaluation type
human preference
Arena text preference / longer_queryhuman preference · 2026-10-15 · rating

Qwen3-30B-A3B-Instruct-2507

1,378.36 rating

Confidence interval: 1,369.83 – 1,386.88

4942 votes

Category
longer_query
Evaluated configuration
qwen3-30b-a3b-instruct-2507
Evaluation type
human preference
Arena text preference / mathhuman preference · 2026-10-15 · rating

Qwen3-30B-A3B-Instruct-2507

1,394.57 rating

Confidence interval: 1,379.45 – 1,409.69

1395 votes

Category
math
Evaluated configuration
qwen3-30b-a3b-instruct-2507
Evaluation type
human preference
Arena text preference / multi_turnhuman preference · 2026-10-15 · rating

Qwen3-30B-A3B-Instruct-2507

1,378.35 rating

Confidence interval: 1,368.82 – 1,387.89

3932 votes

Category
multi_turn
Evaluated configuration
qwen3-30b-a3b-instruct-2507
Evaluation type
human preference
Arena text preference / non_englishhuman preference · 2026-10-15 · rating

Qwen3-30B-A3B-Instruct-2507

1,371.68 rating

Confidence interval: 1,365.83 – 1,377.53

13555 votes

Category
non_english
Evaluated configuration
qwen3-30b-a3b-instruct-2507
Evaluation type
human preference
Arena text preference / overallhuman preference · 2026-10-15 · rating

Qwen3-30B-A3B-Instruct-2507

1,383.83 rating

Confidence interval: 1,378.95 – 1,388.71

23322 votes

Category
overall
Evaluated configuration
qwen3-30b-a3b-instruct-2507
Evaluation type
human preference
Arena text preference / polishhuman preference · 2026-10-15 · rating

Qwen3-30B-A3B-Instruct-2507

1,353.19 rating

Confidence interval: 1,337.98 – 1,368.40

1541 votes

Category
polish
Evaluated configuration
qwen3-30b-a3b-instruct-2507
Evaluation type
human preference
Arena text preference / russianhuman preference · 2026-10-15 · rating

Qwen3-30B-A3B-Instruct-2507

1,369.22 rating

Confidence interval: 1,354.94 – 1,383.50

1564 votes

Category
russian
Evaluated configuration
qwen3-30b-a3b-instruct-2507
Evaluation type
human preference
Arena text preference / spanishhuman preference · 2026-10-15 · rating

Qwen3-30B-A3B-Instruct-2507

1,403.42 rating

Confidence interval: 1,381.06 – 1,425.77

715 votes

Category
spanish
Evaluated configuration
qwen3-30b-a3b-instruct-2507
Evaluation type
human preference

Run it locally

Inference · memory estimate

Qwen3-30B-A3B-Instruct-2507

30.53 B total parameters · qwen3_moe · BF16

Published checkpoint tensor payload: 56.9 GiB. This includes stored copies and buffers; it is separate from the runtime estimates below.

Estimated accelerator memory17.6 GiB
Weights
15.6 GiB
KV cache
0.4 GiB
Runtime allowance
1.6 GiB

One user, full model in accelerator memory.

GPUs with memory headroom

Local AI systems with memory headroom

Shared-memory systems use a planning budget of 75% of installed memory, leaving 25% for the OS and other allocations. This is an assumption, not a measured GPU allocation. Firmware and driver limits may be lower and may require configuration. Memory matches do not verify a runtime, quantized model artifact or inference speed.

Includes FP16 KV cache, prompt plus generated tokens, quantization metadata and a runtime allowance. GPU matches reserve 10% of VRAM. Memory fit is not a speed or software-compatibility guarantee; quantized weights and runtime support must be available.

API cost

USD / month

Standard text-token estimate. Excludes taxes, caching, batch discounts, reasoning and multimodal surcharges. Provider limits and regional availability may differ.

Qwen3-30B-A3B-Instruct-2507

Capabilities

Context window
262,144 tokens
Maximum output
16,384 tokens
Reasoning
No
Tool calling
Yes
Open weights
Yes
Model license
apache-2.0
Total parameters
30.53 B
Checkpoint tensor payload
56.87 GiB

Input modalities

  • text

Output modalities

  • text

Processing & platform

Architecture
qwen3_moe