Qwen3.8-Flash-Next
Alibaba · 1 configuration
Intelligence39.8
Input / 1M$0.15
Output / 1M$0.47
Context256K
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 · 0 without both values
Selected: thinking thinking · Click a point to switch configuration
Open-weight experimental preview of the Qwen4 architecture: hybrid-attention MoE (125B total, 6B active) with vision encoder for coding, agent tasks, and image and video understanding
Intelligence IndexArtificial Analysis composite index, not a percentage. Different versions of the index are not directly comparable.
39.8 pointsGPQA DiamondGraduate-level scientific questions. Measures specialist reasoning rather than web-search quality.
92.3 %Humanity’s Last ExamDifficult questions across many disciplines, compared under the same evaluation setup.
38.0 %SciCodeCode generation for scientific research problems.
50.6 %Terminal-Bench 2.1Agent tasks in a terminal environment. Results also depend on the evaluation harness.
86.1 %MMMU-ProMultimodal understanding of academic problems that include images.
79.8 %Omniscience accuracyFactual accuracy in Artificial Analysis's Omniscience evaluation.
24.5 %Specifications from the catalog
Inputtext, image, video
Outputtext
Tool callingYes
Structured outputYes
Speed58.1 tok/s
First token2.50 s
Maximum output64K
Catalog listing date2026-08-27