AlibabaAuto-detected

qwen3-4b-instruct-2507

Name: qwen3-4b-instruct-2507 benchmark results
Creator: BenchLeader

qwen3-4b-instruct-2507 is an Alibaba proprietary model, released 6 Aug 2025. Its best configuration ranks #501 of 372 on the BenchLeader Index at 41.4 ±3.8, in the lower half. It scores highest in maths (47) and lowest in long context (31). It has been measured at 2 reasoning-effort settings; tables show the best-scoring one. Last measured 6 Aug 2025.stale: no new result in six months

Blended price: –
Output speed: –
First answer: –
Context: 262k

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

Overall index41
Reasoning36
Agents & tools44
Maths47
Knowledge35
Instruction following37
Long context31
Composite38

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. “Default” means the publisher did not say which setting was used.

Effort	Index	Rank	Speed	First answer	Chat reply cost	Categories
thinking	40.2	#525	–	–	–	Agents & tools 35 · Composite 40 · Instruction following 51 · Knowledge 36 · Long context 45 · Maths 13
defaultbest	41.4	#501	–	–	–	Agents & tools 44 · Composite 38 · Instruction following 37 · Knowledge 35 · Long context 31 · Maths 47 · Reasoning 36

Benchmark results

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	default	Source
GPQA Diamond	–	45.8%#224	Epoch AI Benchmarking Hub
GPQA Diamond (AA)not in index	66.7%#319	51.7%#404	Artificial Analysis
Humanity's Last Exam (AA)not in index	6.2%#339	4.5%#428	Artificial Analysis

Agents & tools

Benchmark	thinking	default	Source
Terminal-Bench Hard	1.5%#328	4.5%#275	Artificial Analysis
τ²-Bench Telecom (AA)not in index	25.4%#283	26.6%#272	Artificial Analysis
BFCL Overall	–	35.7%#41	Berkeley Function Calling Leaderboard

Maths

Benchmark	thinking	default	Source
OTIS Mock AIME	–	52.2%#167	Epoch AI Benchmarking Hub
AIME 2026	82.5%#30	–	MathArena
HMMT February 2026	53.0%#31	–	MathArena
MathArena Apex	2.1%#29	–	MathArena

Knowledge

Benchmark	thinking	default	Source
AA-Omniscience	-60.3#401	-61.9#409	Artificial Analysis

Instruction following

Benchmark	thinking	default	Source
IFBench	49.8%#174	33.5%#315	Artificial Analysis

Long context

Benchmark	thinking	default	Source
AA-LCR	37.3%#301	11.3%#394	Artificial Analysis

Composite

Benchmark	thinking	default	Source
AA Intelligence Index	8.8#356	6.7#437	Artificial Analysis