What are Qwen3.6 35B A3B's benchmark scores?

As of 24 Oct 2026, Qwen3.6 35B A3B (thinking effort) scores 52.2 ±6.6 on the BenchLeader Index, strongest in instruction following (64) and weakest in reasoning (41). That rests on 8 independent benchmark results, and every score on this page links to its publisher.

How does Qwen3.6 35B A3B rank?

Qwen3.6 35B A3B (thinking effort) is #327 of 785 ranked configurations on 24 Oct 2026. 372 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 72.6.

How much does Qwen3.6 35B A3B cost per 1M tokens?

As of 24 Oct 2026, Qwen3.6 35B A3B costs $0.380 per 1M input tokens and $2.25 per 1M output tokens, $0.848 blended (three input tokens per output token). The list price last changed on 6 Oct 2026.

When was Qwen3.6 35B A3B released?

Alibaba released Qwen3.6 35B A3B on 14 Apr 2026. BenchLeader has results for it at 3 reasoning-effort settings.

Alibaba Open weights: Qwen/Qwen3.6-35B-A3BReasoning modelCompare with another model

Qwen3.6 35B A3B

Reasoning effort

Qwen3.6 35B A3B is an Alibaba open-weights reasoning model, released 14 Apr 2026. Its best configuration (thinking reasoning effort) ranks #327 of 785 on the BenchLeader Index at 52.2 ±6.6, in the upper half. The ± is the point: 372 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 instruction following (64) and lowest in reasoning (41). At $0.848 per million tokens blended it is mid-priced. Output speed of 136 tokens per second puts it faster than most, with a first answer in 41.5 s. It has been measured at 3 reasoning-effort settings; this summary describes the best-scoring one, and the tabs above switch between them. Last measured 16 Apr 2026.stale: no new result in six months

Blended price: $0.848/M; $0.380 in · $2.25 out
Output speed: 136 tok/s; measured by Artificial Analysis
First answer: 42 s; first token 0.63 s
Context: 262k
Full answer: 45 s; median, reasoning included
Cost per run: $0.475; one full Intelligence Index run
Released: 14 Apr 2026

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

Overall index52
Reasoning41
Agents & tools46
Knowledge52
Instruction following64
Multimodal58
Long context61
Composite50

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
Qwen3.6 35B A3Bthinkingthis page	14 Apr 2026	52.2	#327
Qwen3.5-35B-A3Bthinking	24 Feb 2026	52.8	#305

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
no reasoning	42.9	#581	155 tok/s	1.99 s	$0.0008	Agents & tools 41 · Coding 5 · Composite 47 · Instruction following 39 · Knowledge 34 · Long context 57 · Maths 42 · Multimodal 54 · Reasoning 51
thinkingbest	52.2	#327	136 tok/s	42 s	$0.0008	Agents & tools 46 · Composite 50 · Instruction following 64 · Knowledge 52 · Long context 61 · Multimodal 58 · Reasoning 41
not stated	46.3	#493	71 tok/s	0.63 s	$0.0008	Agents & tools 36 · Coding 37 · Maths 44 · Reasoning 61

How its index has moved

8 Sept 2026 to 21 Oct 2026

thinking
not stated
no reasoning

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	no reasoning	thinking	not stated	Source
GPQA Diamond	84.8%#93	–	83.8%#99	Epoch AI Benchmarking Hub
GPQA Diamond (AA)not in index	81.7%#178	84.1%#149	–	Artificial Analysis
Humanity's Last Exam (AA)not in index	13.9%#264	22.2%#201	–	Artificial Analysis
CritPt	0.0%#315	0.3%#271	–	Artificial Analysis
Mystery Game Puzzlesnot in index	22.0%#55	–	–	Epoch AI Benchmarking Hub
LMCAnot in indexeffort not stated	–	–	29.7%#135	Epoch AI Benchmarking Hub
DTBenchnot in indexeffort not stated	–	–	73.9%#152	Epoch AI Benchmarking Hub

Coding

Benchmark	no reasoning	thinking	not stated	Source
SciCode	1.3%#216	–	36.6%#181	SciCode
WeirdMLeffort not stated	–	–	34.5%#129	WeirdML
SciCode (AA)not in index	–	36.6%#193	–	Artificial Analysis
Surface Evolver Benchnot in indexeffort not stated	–	–	44.4%#22	Epoch AI Benchmarking Hub
Terminal-Bench 4.0 (AA)not in index	–	0.0%#152	–	Artificial Analysis
Terminal-Bench 2.1 (AA)not in index	41.6%#136	44.9%#128	–	Artificial Analysis

Agents & tools

Benchmark	no reasoning	thinking	not stated	Source
Terminal-Bencheffort not stated	–	–	24.6%#66	Terminal-Bench
Terminal-Bench Hard	25.8%#135	34.9%#82	–	Artificial Analysis
τ²-Bench Telecom (AA)not in index	85.1%#94	95.3%#23	–	Artificial Analysis
GDPval-AA v2.1not in index	17.3%#102	19.6%#99	–	Artificial Analysis
τ³-Banking (AA)	5.4%#179	9.3%#150	–	Artificial Analysis
AutomationBenchnot in index	–	5.2%#164	–	Artificial Analysis
GDP.pdf (AA)not in index	–	6.6%#150	–	Artificial Analysis

Maths

Benchmark	no reasoning	not stated	Source	Trend
FrontierMath Tiers 1–3	20.4%#87	17.5%#98	Epoch AI Benchmarking Hub
OTIS Mock AIME	68.9%#141	86.7%#86	Epoch AI Benchmarking Hub

Knowledge

Benchmark	no reasoning	thinking	Source
AA-Omniscience	-60.0#465	-22.2#249	Artificial Analysis
AA-Omniscience: accuracynot in index	16.7%#382	18.8%#327	Artificial Analysis
AA-Omniscience: non-hallucinationnot in index	7.9%#474	49.5%#118	Artificial Analysis

Instruction following

Benchmark	no reasoning	thinking	Source	Trend
IFBench	36.2%#304	64.3%#110	Artificial Analysis

Multimodal

Benchmark	no reasoning	thinking	Source	Trend
MMMU-Pro	71.0%#140	75.0%#104	Artificial Analysis

Long context

Benchmark	no reasoning	thinking	Source	Trend
AA-LCR	64.3%#257	71.7%#196	Artificial Analysis
MLCRnot in index	–	5.0%#144	Artificial Analysis

Composite

Benchmark	no reasoning	thinking	not stated	Source	Trend
Epoch Capabilities Indexnot in indexeffort not stated	–	–	143.7#93	Epoch AI Benchmarking Hub
AA Intelligence Index v4.3.2	15.2#286	18.2#256	–	Artificial Analysis

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
AtlasCloud	153 tok/s	2.23 s	$0.250	$1.25	262k	fp8
CoreWeave	127 tok/s	0.45 s	$0.250	$1.25	262k	fp8
Venice	115 tok/s	0.78 s	$0.100	$1.00	256k	fp8
DekaLLM	112 tok/s	0.80 s	$0.100	$1.00	262k	–
Parasail	92 tok/s	0.47 s	$0.100	$0.950	262k	fp8
DeepInfra	83 tok/s	0.73 s	$0.100	$0.950	262k	fp8
Phala	66 tok/s	1.18 s	$0.200	$1.27	262k	–
AkashML	63 tok/s	0.60 s	$0.100	$0.900	262k	fp8
Darkbloom	43 tok/s	3.12 s	$0.050	$0.700	262k	fp4
SiliconFlow	31 tok/s	1.82 s	$0.240	$1.80	262k	fp8

Price history

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

input outputnow $0.380 in · $2.25 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	$0.0008	$0.0007	43.7 s
Summarize a 30-page report	12,000 / 600	$0.0059	$0.0034	45.9 s
Code edit	6,000 / 1,500	$0.0057	$0.0044	52.5 s
Agentic coding session	60,000 / 4,000	$0.032	$0.019	1.2 min
Structured extraction	2,000 / 200	$0.0012	$0.0008	43.0 s

New this month

Questions people ask

What are Qwen3.6 35B A3B's benchmark scores?: As of 24 Oct 2026, Qwen3.6 35B A3B (thinking effort) scores 52.2 ±6.6 on the BenchLeader Index, strongest in instruction following (64) and weakest in reasoning (41). That rests on 8 independent benchmark results, and every score on this page links to its publisher.
How does Qwen3.6 35B A3B rank?: Qwen3.6 35B A3B (thinking effort) is #327 of 785 ranked configurations on 24 Oct 2026. 372 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 72.6.
How much does Qwen3.6 35B A3B cost per 1M tokens?: As of 24 Oct 2026, Qwen3.6 35B A3B costs $0.380 per 1M input tokens and $2.25 per 1M output tokens, $0.848 blended (three input tokens per output token). The list price last changed on 6 Oct 2026.
When was Qwen3.6 35B A3B released?: Alibaba released Qwen3.6 35B A3B on 14 Apr 2026. BenchLeader has results for it at 3 reasoning-effort settings.

Data as of 24 Oct 2026. Compare these configurations.

Cite as: BenchLeader, “Qwen3.6 35B A3B: benchmarks, pricing, speed and rank”, data as of 24 Oct 2026, https://www.benchleader.com/models/qwen3-6-35b-a3b.