Qwen3 30B A3B 2507
Alibaba · 2 thinking configurations
Intelligence9.8
Input / 1M$0.20
Output / 1M$2.40
Context262.1K
Tariff: Alibaba · Source: artificial-analysis · USD per million tokens
Thinking configuration
What does more thinking buy you?
Compare this model’s measured thinking configurations.
Scroll the chart ↔ · tap a point for details
Efficient frontier: best measured score at each cost1 measured configurations · 1 without both values
Selected: thinking thinking · Click a point to switch configuration
Qwen instruction model for multilingual chat, reasoning, and tool use
Intelligence IndexArtificial Analysis composite index, not a percentage. Different versions of the index are not directly comparable.
9.8 pointsGPQA DiamondGraduate-level scientific questions. Measures specialist reasoning rather than web-search quality.
70.7 %Humanity’s Last ExamDifficult questions across many disciplines, compared under the same evaluation setup.
10.3 %SciCodeCode generation for scientific research problems.
33.0 %Terminal-Bench 2.1Agent tasks in a terminal environment. Results also depend on the evaluation harness.
1.5 %τ²-BenchTool use and interaction in assistance scenarios. Does not replace an evaluation of your integration.
28.1 %IFBenchVerifiable instruction following. A proxy for precision rather than a benchmark of creative style.
50.7 %Omniscience accuracyFactual accuracy in Artificial Analysis's Omniscience evaluation.
16.3 %Specifications from the catalog
Inputtext
Outputtext
Tool callingYes
Structured outputNot available
Speed134.3 tok/s
First token2.40 s
Maximum output16.4K
Catalog listing date2025-07-30