Skip to content

Alibaba · Chat

Qwen3 30B A3B Instruct

Alibaba 🇨🇳ChatQwen familyOpen weights. The weights are publicly available in a Hugging Face repo.
From Eurouter's listing:
Qwen3 30B A3B Instruct is an Alibaba/Qwen compact MoE instruction model for chat, code, tool use, and context-augmented generation.
llminfo IQ
114.9
Market median price
$0.15 blended / 1M
Max context
262K
Listings
27

Evidence and sources

International storefront (alibabacloud.com) is Singapore-incorporated; mainland operating entity 杭州阿里云智能科技有限公司 (Hangzhou).

🤗 Qwen/Qwen3-30B-A3B-Instruct-2507llminfo IQ 114.9, ranked among chat models. From 100 benchmarks. Plenty of data behind it.How llminfo IQ is computedCoding IQ 86.4, 95% interval 67–106. From 3 benchmark families across 3 sources. Little data behind it, so treat it as a rough estimate. Uses the model's best published results. It has its own scale, separate from llminfo IQ.How Coding IQ is computedPoor value (36.5). Costs more per task than its benchmark results predict. From 5 benchmarks with measured task costs. A fair amount of data behind it.How Value for money worksReleased 28 Jul 2025Maker HQ: CN (source)
All rankings: #160 on OpenRouter this week #161 on OpenRouter today #172 on OpenRouter this month #426 on NanoGPT recentlySources: LMArena data, EURouter data, Hugging FacePrimary data from LMArena from EURouter: name, description from Hugging Face: release date
Tags from its sources:open-sourcemoelong-contextcodingfunction-calling

Advantages and trade-offs

Compared with 10 chat models, picked mainly by popularity and IQ

Advantages

Open weights available

The weights are public. 2 of the 7 peers that say either way have open weights.

We only make a claim when there's enough data on models of the same type. Updated 22 Nov, 06:07 UTC.Prices come from the model's maker where they sell it, otherwise their official partner, otherwise the market median. This comparison uses 7 maker's prices; 3 market median prices.

Input & output

as stated by the model's providers
Text
Text

Capabilities

Supported operations and provider-stated features

ReasoningFunction callingStructured outputPrompt caching

Listings

 Flowdotqwen/qwen3-30b-a3b-instruct-2507Source: FlowDot data262K$0.09$0.30——
 Hugging FaceQwen/Qwen3-30B-A3B-Instruct-2507Source: Models.dev data262K$0.09$0.30——
 OpenRouterInference by SiliconFlowqwen/qwen3-30b-a3b-instruct-2507 · fp8Source: OpenRouter data docs262K$0.09$0.30171,015ms
 Ambientqwen/qwen3-30b-a3b-instruct-2507Sources: Ambient data, Models.dev data131K$0.09$0.30——
 SiliconFlowQwen/Qwen3-30B-A3B-Instruct-2507Sources: Models.dev data, SiliconFlow data262K$0.09$0.30——
 Cortecs AIqwen3-30b-a3b-instruct-2507Sources: Cortecs EU Router data, Models.dev data262K——
 302aiQwen/Qwen3-30B-A3B-Instruct-2507Source: 302.ai data256K$0.10$0.40——
 Eden AIInference by Nebius AI Studionebius/Qwen/Qwen3-30B-A3B-Instruct-2507Source: Models.dev data262K$0.10$0.30——
 Kiloqwen/qwen3-30b-a3b-instruct-2507Sources: Kilo AI Gateway data, Models.dev data262K$0.10$0.30——
 Nous Portalqwen/qwen3-30b-a3b-instruct-2507Source: Nous Portal Inference Gateway data262K$0.10$0.30——
Show all 23 listings

Showing 10 of 23: the cheapest per 1M input tokens from each provider.

Also sold by Aihubmix, Eurouter, Fireworks AI, NanoGPT, Nebius AI Studio, Prime, Qiniu, Requesty, Siliconflow Cn, Weights & Biases.

4 listings without a price are hidden.

Benchmarks

Named effort levels run from lowest to highest. Select a result to see its source details.

Human preference

Arena ELO27 related benchmarks · 27 results

Multilingual

Benchmark Nusantara19 related benchmarks · 19 results
id-N terbuka (prompt open-py1, scorer open-py1, items 5ca9dfe11dc2, key eda1d1cddc75, public slice)
Source: Benchmark Nusantara (ibahasa) data docs · Aug 2026
repetition: 1, route qwen/qwen3-30b-a3b-instruct-2507, token cap 600
id-B pilihan-ganda (prompt mcq-py1, scorer mcq-py1, items 3e5973f02ae8, key 4e94e1cb7f25, public slice)
Source: Benchmark Nusantara (ibahasa) data docs · Aug 2026
repetition: 1, route qwen/qwen3-30b-a3b-instruct-2507, token cap 600
id-B pilihan-ganda (prompt mcq-py1, scorer mcq-py1, items 3e5973f02ae8, key 4e94e1cb7f25, withheld slice)
Source: Benchmark Nusantara (ibahasa) data docs · Aug 2026
repetition: 1, route qwen/qwen3-30b-a3b-instruct-2507, token cap 600
id-B pilihan-ganda (prompt mcq-py1, scorer mcq-py1, items d3e8c01bc554, key 4e94e1cb7f25, public slice)
Source: Benchmark Nusantara (ibahasa) data docs · Aug 2026
repetition: 1, route qwen/qwen3-30b-a3b-instruct-2507, token cap 600
id-B pilihan-ganda (prompt mcq-py1, scorer mcq-py1, items d3e8c01bc554, key 4e94e1cb7f25, withheld slice)
Source: Benchmark Nusantara (ibahasa) data docs · Aug 2026
repetition: 1, route qwen/qwen3-30b-a3b-instruct-2507, token cap 600
id-N terbuka (prompt open-py1, scorer open-py1, items 5ca9dfe11dc2, key eda1d1cddc75, withheld slice)
Source: Benchmark Nusantara (ibahasa) data docs · Aug 2026
repetition: 1, route qwen/qwen3-30b-a3b-instruct-2507, token cap 600
Show 13 more resultsHide 13 results
id-N terbuka (prompt open-py2, scorer open-py1, items 5ca9dfe11dc2, key eda1d1cddc75, public slice)
Source: Benchmark Nusantara (ibahasa) data docs · Aug 2026
repetition: 1, route qwen/qwen3-30b-a3b-instruct-2507, token cap 600
id-N terbuka (prompt open-py2, scorer open-py1, items 5ca9dfe11dc2, key eda1d1cddc75, withheld slice)
Source: Benchmark Nusantara (ibahasa) data docs · Aug 2026
repetition: 1, route qwen/qwen3-30b-a3b-instruct-2507, token cap 600
id-N terbuka (prompt open-py3, scorer open-py1, items 5ca9dfe11dc2, key eda1d1cddc75, public slice)
Source: Benchmark Nusantara (ibahasa) data docs · Aug 2026
repetition: 1, route qwen/qwen3-30b-a3b-instruct-2507, token cap 600
id-N terbuka (prompt open-py3, scorer open-py1, items 5ca9dfe11dc2, key eda1d1cddc75, withheld slice)
Source: Benchmark Nusantara (ibahasa) data docs · Aug 2026
repetition: 1, route qwen/qwen3-30b-a3b-instruct-2507, token cap 600
id-W terbuka (prompt open-py1, scorer open-py1, items 27af2c184856, key 82bbf3cc2f8a, public slice)
Source: Benchmark Nusantara (ibahasa) data docs · Aug 2026
repetition: 1, route qwen/qwen3-30b-a3b-instruct-2507, token cap 600
id-W terbuka (prompt open-py1, scorer open-py1, items 27af2c184856, key 82bbf3cc2f8a, withheld slice)
Source: Benchmark Nusantara (ibahasa) data docs · Aug 2026
repetition: 1, route qwen/qwen3-30b-a3b-instruct-2507, token cap 600
jv-A pilihan-ganda (prompt mcq-py1, scorer mcq-py1, items c0c83850858e, key 18e8ec378904, public slice)
Source: Benchmark Nusantara (ibahasa) data docs · Aug 2026
repetition: 1, route qwen/qwen3-30b-a3b-instruct-2507, token cap 600

General

Science

GPQA Diamond10 results
Source: Epoch AI Benchmarking Hub data
Setting not stated
Source: OpenRouter benchmark exports
qwen3-30b-a3b-instruct-2507 · 90-day results
Across 8 providers lowest 62.1% median 63.3% highest 65.2%OpenRouter's automatic routing 65.2%
Show results from 8 providers and OpenRouter's automatic routingHide results from 8 providers and OpenRouter's automatic routing
  • OpenRouter's automatic routing
  • Alibaba
  • Ambient
  • CoreWeave
  • SiliconFlow
  • StreamLake
  • DekaLLM
  • PrimeIntellect
  • Nebius

Coding

SWE-rebenchSource: SWE-rebench

Agentic

Source: OpenRouter benchmark exports
qwen3-30b-a3b-instruct-2507 · 90-day results
Across 7 providers lowest 35% median 38.7% highest 42.2%OpenRouter's automatic routing 38.3%
Show results from 7 providers and OpenRouter's automatic routingHide results from 7 providers and OpenRouter's automatic routing
  • StreamLake
  • SiliconFlow
  • Ambient
  • CoreWeave
  • OpenRouter's automatic routing
  • PrimeIntellect
  • Nebius
  • Alibaba

Math

Reasoning

Safety

PhareSource: Phare

Instruction following

Factuality

Other

Data credits: LMArena, CC BY 4.0 (licence) · ArtifactsBench (Tencent Hunyuan), CC BY 4.0 (licence) · Benchmark Nusantara, ibahasa (M. Khoirul Huda), doi:10.5281/zenodo.21970463. Test results CC BY 4.0, attribution to ibahasa; aggregate results CSV only. (licence) · Epoch AI Benchmarking Hub, CC BY 4.0 (licence) · OpenRouter public benchmark leaderboard and search-lane exports, CC BY 4.0. Fractional source scores converted to percentages; immutable dated snapshots. (licence) · EQ-Bench (Samuel J. Paech) — cite paech2023eqbench · Open Arabic LLM Leaderboard (OALL), Apache-2.0 (licence) · OpenLLM French Leaderboard (le-leadboard), Mohamad Alhajar and Alexandre Lavallée; publisher-normalized public result files; results licence not stated (the Space's Apache-2.0 covers code) · Phare (Giskard), CC BY 4.0 (licence) · Shaberi v2.1, Shisa.AI, Apache-2.0. GPT-5.1-judge scores only; judges are not comparable. (licence) · SnakeBench (Greg Kamradt / ARC Prize Foundation), MIT (licence)

Also known as: fireworks_ai/accounts/fireworks/models/qwen3-30b-a3b-instruct-2507, nebius/qwen/qwen3-30b-a3b-instruct-2507, qwen/qwen3-30b-a3b-instruct-2507, qwen3-30b-a3b-instruct-2507, sophnet/qwen3-30b-a3b-instruct-2507, wandb/qwen/qwen3-30b-a3b-instruct-2507

Like this one. Tap or hover any dotted value, label or icon to see where it comes from and how it is measured. open details when tapped or hovered. How this data works