What are Qwen3 30B A3B 2507's benchmark scores?

As of 2 Nov 2026, Qwen3 30B A3B 2507 scores 48.1 ±5.0 on the BenchLeader Index, strongest in coding (56) and weakest in instruction following (37). That rests on independent benchmark results only, and every score on this page links to its publisher.

How does Qwen3 30B A3B 2507 rank?

Qwen3 30B A3B 2507 is #376 of 585 ranked configurations on 2 Nov 2026. 258 other configurations score within its margin, so the exact position matters less than the band. The leader is Claude Opus 5.5 (max effort) at 69.2.

How much does Qwen3 30B A3B 2507 cost per 1M tokens?

As of 2 Nov 2026, Qwen3 30B A3B 2507 costs $0.100 per 1M input tokens and $0.300 per 1M output tokens, $0.150 blended (three input tokens per output token). The list price last changed on 21 Oct 2026.

When was Qwen3 30B A3B 2507 released?

Alibaba released Qwen3 30B A3B 2507 on 30 Jul 2025. BenchLeader has results for it at 2 reasoning-effort settings.

Alibaba Open weights: Qwen/Qwen3-30B-A3B-Instruct-2507Reasoning modelCompare with another model

Qwen3 30B A3B 2507

Reasoning effort

Quality#186 of 329

48.1Index

Below the median

−2.0 on the median

Price#23 of 246

$0.150per 1M tokens

Among the cheapest 10%

0.2× the median

Speed#25 of 210

201tokens/s

Top quarter

2.4× the median

4 cheaper ranked models score higher: GPT-5 nano (55.9, $0.138), Qwen3.5 Flash (51.9, $0.094), Ling 3.0 Flash (51.8, $0.115) and 1 more.

As of 2 Nov 2026, Qwen3 30B A3B 2507 by Alibaba scores 48.1 on the BenchLeader Index (585 ranked configurations), with its best results in coding (56) and weakest in instruction following (37). API pricing: $0.100 per 1M input tokens and $0.300 per 1M output tokens ($0.150 blended at 3:1), list price on 2 Nov 2026. It streams about 200 tokens/s (Artificial Analysis). Released 30 Jul 2025.

Key facts (as of 2 Nov 2026)

Qwen3 30B A3B 2507 API pricing: $0.100 per 1M input tokens and $0.300 per 1M output tokens ($0.150 blended at 3:1) (list price, 2 Nov 2026).
Qwen3 30B A3B 2507 streams about 200 tokens/s (Artificial Analysis, 2 Nov 2026).
Qwen3 30B A3B 2507 was released on 30 Jul 2025 (Alibaba).

In detail. Qwen3 30B A3B 2507 is an Alibaba open-weights reasoning model, released 30 Jul 2025. Its best configuration ranks #376 of 585 on the BenchLeader Index at 48.1 ±5.0, in the lower half. The ± is the point: 258 other configurations score within that range, so they and this one cannot be told apart on quality alone — price and speed are what separate them. It scores highest in coding (56) and lowest in instruction following (37). At $0.150 per million tokens blended it is among the cheapest fifth of ranked models. Output speed of 201 tokens per second puts it in the fastest quarter, with a first answer in 0.9 s. It has been measured at 2 reasoning-effort settings; this summary describes the best-scoring one, and the tabs above switch between them. Last measured 28 Oct 2026.

Blended price: $0.150/M; $0.100 in · $0.300 out; via Artificial Analysis
Output speed: 201 tok/s; measured by Artificial Analysis
First answer: 0.93 s; first token 0.30 s
Context: 262k; via Artificial Analysis
Full answer: 3.42 s; median, reasoning included
Cost per task: –
Tokens per task: –
Released: 30 Jul 2025

How it scores by categoryDashed line = average model (50). One step of 15 = one standard deviation.

Overall index48
Reasoning46
Coding56
Agents & tools48
Maths53
Instruction following37
Human preference55
Long context37

Versions

Alibaba has shipped 2 models under this name. Each is ranked on its own results; a newer version often has fewer results so far, which holds its index nearer the average until more arrive.

Model	Released	Index	Rank
Qwen3 30B A3B 2507this page	30 Jul 2025	48.1	#376
Qwen3 30B A3B	29 Apr 2025	48.2	#375

Reasoning-effort configurations

The same model behaves differently depending on how much it is allowed to think. Each row is one setting, scored only on the benchmarks that were run at that setting. “Not stated” collects results from publishers that did not say which setting they used; for a reasoning model that is usually its thinking mode, but we do not assume it. Pick a setting here or at the top of the page to see its price, speed and category scores.

Effort	Index	Rank	Speed	First answer	Chat reply cost	Categories
thinking	44.8	#445	201 tok/s	0.93 s	$0.0001	Agents & tools 35 · Coding 30 · Instruction following 52 · Long context 55 · Maths 38 · Reasoning 51
not statedbest	48.1	#376	201 tok/s	0.93 s	$0.0001	Agents & tools 48 · Coding 56 · Human preference 55 · Instruction following 37 · Long context 37 · Maths 53 · Reasoning 46

How its index has moved

8 Sep 2026 to 29 Oct 2026

not stated
thinking

The index is recomputed from scratch every day, so a line moves when a new benchmark result lands, when a publisher revises a score, or when the models it is normalised against change. Early movement usually means the score is still settling.

Benchmark results

See all LLM benchmarks. One column per reasoning effort. Rank is among every configuration of every model on that benchmark. Hover a score for the run it came from.

Reasoning

Benchmark	thinking	not stated	Source
GPQA Diamond	70.1%#173	55.6%#205	Epoch AI Benchmarking Hub
Arena Hard Promptseffort not stated	–	1407#170	Arena (LMArena)
GPQA Diamond (AA)not in index	70.7%#297	65.9%#341	Artificial Analysis
Humanity's Last Exam (AA)not in index	10.3%#329	6.9%#388	Artificial Analysis
Kagi LLM Benchmark	54.9%#67	–	Kagi LLM Benchmark
CritPtnot in index	0.3%#282	0.0%#328	Artificial Analysis
LMCAnot in index	19.5%#160	22.4%#155	Epoch AI Benchmarking Hub
DTBenchnot in index	69.3%#158	67.2%#167	Epoch AI Benchmarking Hub

Coding

Benchmark	thinking	not stated	Source
SciCode	33.0%#204	–	SciCode
Arena Codingeffort not stated	–	1438#161	Arena (LMArena)
SciCode (AA)not in index	33.0%#212	–	Artificial Analysis
Terminal-Bench 4.0 (AA)not in index	0.0%#167	–	Artificial Analysis
Terminal-Bench 2.1 (AA)not in index	1.5%#224	–	Artificial Analysis
LiveCodeBench (AA)not in index	70.7%#62	51.5%#134	Artificial Analysis

Agents & tools

Benchmark	thinking	not stated	Source
Terminal-Bench Hard	5.3%#279	6.1%#268	Artificial Analysis
τ²-Bench Telecom (AA)not in index	28.1%#271	10.2%#388	Artificial Analysis
BFCL Overalleffort not stated	–	41.4%#33	Berkeley Function Calling Leaderboard
τ³-Banking (AA)	5.4%#181	–	Artificial Analysis
AutomationBenchnot in index	0.8%#215	–	Artificial Analysis
GDP.pdf (AA)not in index	0.4%#231	–	Artificial Analysis

Maths

Benchmark	thinking	not stated	Source
OTIS Mock AIME	70.3%#141	62.2%#161	Epoch AI Benchmarking Hub
AIME 2026	88.3%#29	–	MathArena
HMMT February 2026	78.8%#27	–	MathArena
MathArena Apex	0.5%#41	–	MathArena
Arena Mathseffort not stated	–	1378#177	Arena (LMArena)

Instruction following

Benchmark	thinking	not stated	Source	Trend
IFBench	50.7%#172	33.1%#336	Artificial Analysis
Arena Instruction Followingnot in indexeffort not stated	–	1366#180	Arena (LMArena)

Human preference

Benchmark	not stated	Source
Arena Text	1382#176	Arena (LMArena)
Arena Creative Writingnot in index	1318#199	Arena (LMArena)
Arena Multi-turnnot in index	1379#176	Arena (LMArena)
Arena Longer Queriesnot in index	1381#178	Arena (LMArena)

Long context

Benchmark	thinking	not stated	Source	Trend
AA-LCR v1.1	61.3%#283	26.3%#419	Artificial Analysis
MLCRnot in index	2.2%#163	–	Artificial Analysis

Composite

Benchmark	thinking	not stated	Source
Epoch Capabilities Indexnot in index	139.6#117	137.3#128	Epoch AI Benchmarking Hub
AA Intelligence Index v4.3.2not in index	9.8#400	7.5#480	Artificial Analysis
AA Progress Index v1not in index	214.2#368	198.5#420	Artificial Analysis

API pricing: where to run it

Every provider serving this model through OpenRouter, with throughput and first-token latency measured on live traffic over the last 30 minutes and each provider’s own price. Purple marks the best in each column.

Provider	Speed	First token	Input $/M	Output $/M	Context	Quantisation
Nebius Token Factory	161 tok/s	0.23 s	$0.100	$0.300	262k	fp8
DekaLLM	72 tok/s	0.46 s	$0.090	$0.300	262k	–
SiliconFlow	28 tok/s	1.01 s	$0.090	$0.300	262k	fp8

Price history

The price above, per 1M tokens, every day it changed.

input outputnow $0.100 in · $0.300 out

What a task costs

Estimates from list price, output speed and time to first answer for the best configuration. “With caching” assumes three-quarters of the input is served from the prompt cache at $0.100 per 1M. Reasoning tokens are not modelled.

Workload	Tokens in / out	Cost	With caching	Time
Chat reply	400 / 300	$0.0001	$0.0001	2.4 s
Summarize a 30-page report	12,000 / 600	$0.0014	$0.0014	3.9 s
Code edit	6,000 / 1,500	$0.0011	$0.0011	8.4 s
Agentic coding session	60,000 / 4,000	$0.0072	$0.0072	20.9 s
Structured extraction	2,000 / 200	$0.0003	$0.0003	1.9 s

New this month

Embed or cite

Show where Qwen3 30B A3B 2507 ranks on a model card or README. The badge redraws itself with every daily update, carries the data date and links back to this page. To see ranks like this on OpenRouter and Hugging Face model pages, get the Chrome extension.

[![Qwen3 30B A3B 2507 on BenchLeader: #182 overall](https://www.benchleader.com/badge/qwen3-30b-a3b-2507.svg)](https://www.benchleader.com/models/qwen3-30b-a3b-2507)

Cite this

The ranking changes daily, so a reference carries the data date. APA style:

BenchLeader. (2026, November 2). Qwen3 30B A3B 2507 on the BenchLeader Index (method v1.4) [Data set]. Retrieved November 2, 2026, from https://www.benchleader.com/models/qwen3-30b-a3b-2507

BibTeX:

@misc{benchleader_qwen3_30b_a3b_2507_20261102,
  author       = {{BenchLeader}},
  title        = {{Qwen3 30B A3B 2507 on the BenchLeader Index (method v1.4)}},
  year         = {2026},
  month        = nov,
  howpublished = {\url{https://www.benchleader.com/models/qwen3-30b-a3b-2507}},
  note         = {Data as of 2026-11-02}
}

Every badge style, the embeddable top-10 widget and how to cite the Index

Questions people ask

What are Qwen3 30B A3B 2507's benchmark scores?: As of 2 Nov 2026, Qwen3 30B A3B 2507 scores 48.1 ±5.0 on the BenchLeader Index, strongest in coding (56) and weakest in instruction following (37). That rests on independent benchmark results only, and every score on this page links to its publisher.
How does Qwen3 30B A3B 2507 rank?: Qwen3 30B A3B 2507 is #376 of 585 ranked configurations on 2 Nov 2026. 258 other configurations score within its margin, so the exact position matters less than the band. The leader is Claude Opus 5.5 (max effort) at 69.2.
How much does Qwen3 30B A3B 2507 cost per 1M tokens?: As of 2 Nov 2026, Qwen3 30B A3B 2507 costs $0.100 per 1M input tokens and $0.300 per 1M output tokens, $0.150 blended (three input tokens per output token). The list price last changed on 21 Oct 2026.
When was Qwen3 30B A3B 2507 released?: Alibaba released Qwen3 30B A3B 2507 on 30 Jul 2025. BenchLeader has results for it at 2 reasoning-effort settings.

Data as of 2 Nov 2026. Compare these configurations.

Cite as: BenchLeader, “Qwen3 30B A3B 2507: benchmarks, pricing, speed and rank”, data as of 2 Nov 2026, https://www.benchleader.com/models/qwen3-30b-a3b-2507.