Hy3
Tencent · 1 configuration
Intelligence25.3
Input / 1M$0.14
Output / 1M$0.55
Context256K
Tariff: Tencent · 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
Hy3 is a 295B-parameter Mixture-of-Experts model from Tencent (21B active, 192 experts with top-8 routing) built for reasoning, agentic workflows, and real-world production use. It supports a configurable reasoning effort:...
Intelligence IndexArtificial Analysis composite index, not a percentage. Different versions of the index are not directly comparable.
25.3 pointsGPQA DiamondGraduate-level scientific questions. Measures specialist reasoning rather than web-search quality.
89.7 %Humanity’s Last ExamDifficult questions across many disciplines, compared under the same evaluation setup.
33.5 %SciCodeCode generation for scientific research problems.
48.6 %Terminal-Bench 2.1Agent tasks in a terminal environment. Results also depend on the evaluation harness.
64.4 %Omniscience accuracyFactual accuracy in Artificial Analysis's Omniscience evaluation.
32.0 %Specifications from the catalog
Inputtext
Outputtext
Tool callingYes
Structured outputYes
Speed87.9 tok/s
First token3.03 s
Maximum output128K
Catalog listing date2026-07-06